Topologies and control strategies of multi-functional grid-connected inverters for power quality enhancement: A comprehensive review

Abstract

Grid-connected inverters are key components of distributed generation systems (DGSs) and micro-grids (MGs), because they are effective interfaces for renewable and sustainable distributed energy resources (DERs). Recently, multi-functional grid-connected inverters (MFGCIs) have attracted more and more attention for their benefits on auxiliary services on power quality enhancement in DGSs and MGs. These kinds of converters can not only achieve the power generation of DERs, but also can perform as power quality conditioners at their grid-connected points. It should be noted that these functionalities are optimally organized in the same device, which can significantly enhance the cost-effective feature of the grid-connected inverter, as well as can decrease the investment and bulk compared with multiple devices with independent functionalities. MFGCIs are especially suitable for DGSs and MGs application due to their good performances and benefits. Topologies and control strategies of MFGCIs are comprehensively reviewed in this paper. Additionally, detailed explanation, comparison, and discussion on MFGCls are achieved. Furthermore, some future research fields on MFGCls are well summarized.

Full text

Topologies and control strategies of multi-functional grid-connected inverters
for power quality enhancement: A comprehensive review
Zheng Zeng
n
, Huan Yang
1
, Rongxiang Zhao
2
, Chong Cheng
3
College of Electrical Engineering, Zhejiang University, 310027 Hangzhou, Zhejiang Province, China
article info
Article history:
Received 20 March 2012
Received in revised form
11 March 2013
Accepted 15 March 2013
Available online 18 April 2013
Keywords:
Multi-functional grid-connected inverter
Topologies and control strategies
Auxiliary services
Power quality enhancement
Distributed generation system and
micro-grid
Review
abstract
Grid-connected inverters are key components of distributed generation systems (DGSs) and micro-grids
(MGs), because they are effective interfaces for renewable and sustainable distributed energy resources
(DERs). Recently, multi-functional grid-connected inverters (MFGCIs) have attracted more and more
attention for their benefits on auxiliary services on power quality enhancement in DGSs and MGs. These
kinds of converters can not only achieve the power generation of DERs, but also can perform as power
quality conditioners at their grid-connected points. It should be noted that these functionalities are
optimally organized in the same device, which can significantly enhance the cost-effective feature of the
grid-connected inverter, as well as can decrease the investment and bulk compared with multiple
devices with independent functionalities. MFGCIs are especially suitable for DGSs and MGs application
due to their good performances and benefits. Topologies and control strategies of MFGCIs are
comprehensively reviewed in this paper. Additionally, detailed explanation, comparison, and discussion
on MFGCIs are achieved. Furthermore, some future research fields on MFGCIs are well summarized.
&2013 Elsevier Ltd. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
2. Some outlines on grid-connected inverters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
3. Power quality of distributed generation systems and micro-grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
4. Multi-functional grid-connected inverters in single-phase system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
5. Multi-functional grid-connected inverters in three-phase system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
6. Analysis and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
7. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
1. Introduction
Several blackouts caused by chain failures [1], as well as the
electric grid splits because of the extreme weather [2,3] threaten
the security and stability of traditional electric power systems.
In addition, the continuous consumption of fossil fuels is leading to
energy crisis and increasing environmental pollution problems.
Therefore, the “green”and “low carbon”power becomes the urgent
need of traditional electric power systems [4,5]. Facing to these
issues, distributed generation systems (DGSs) gradually return to
the stage [6–8]. Numerous studies show that DGSs can not only
connect renewable energy sources (RESs), such as wind, solar and
so on, to utility grid, but also can improve the stability of traditional
electric power systems in some sense. In order to make better use of
RESs, micro-grids (MGs) considered as special DGSs have been
widely discussed and demonstrated, and are given great expecta-
tions. A micro-grid is a local power supply system, which integrates
RESs, energy storage devices, local loads, communication devices,
protection units, and the control center [9–14]. Recently, DGSs and
MGs are very active and encouraging research fields.
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/rser
Renewable and Sustainable Energy Reviews
1364-0321/$- see front matter &2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.rser.2013.03.033
n
Corresponding author. Tel.: +86 13567123512.
E-mail addresses: zengerzheng@zju.edu.cn (Z. Zeng),
yanghuan@zju.edu.cn (H. Yang), rongxiang@zju.edu.cn (R. Zhao),
chengchong@zju.edu.cn (C. Cheng).
1
Tel.: +86 13588846066.
2
Tel.: +86 13906508946.
3
Tel.: +86 15957136558.
Renewable and Sustainable Energy Reviews 24 (2013) 223–270

In DGSs and MGs, the grid-connected inverters (GCIs) are
essential interfaces to connect RESs and energy storage devices
to utility grid [15,16]. To reduce the investment, operation and
maintenance cost, man-hour, as well as the bulk, and enhance the
cost-effective feature of the GCIs in DGSs and MGs, the multi-
functional grid-connected inverters (MFGCIs) are proposed in
[17–26]. The so-called MFGCIs can connect RESs and storage
devices to utility grid, and simultaneously enhance the power
quality at their points of common coupling (PCCs). Compared with
the multiple devices with different functionalities, the MFGCIs can
greatly save capital investment and system space, because the
different functionalities of multiple devices are integrated in the
same equipment. So MFGCIs are good choices for DGSs and MGs
application and are paid common attention.
In this paper, a comprehensive review on the topologies and
control strategies of MFGCIs is achieved. Meanwhile, from the
views of single-phase and three-phase utility grid, detailed expla-
nation, comparison, and discussion of the MFGCIs are summarized.
Besides, some future frameworks on MFGCIs are presented.
The paper is organized as follows. In Section 2, some important
outlines on GCIs are briefly introduced. The power quality of DGSs
and MGs, as well as some possible response strategies to enhance
the power quality are described in Section 3.InSections 4 and 5,
the available topologies and control strategies of MFGCIs are
comprehensively reviewed for single-phase and three-phase uti-
lity application, respectively. A detailed analysis and comparison of
the available MFGCIs are investigated in Section 6. In addition,
some interesting research points are presented. Some conclusions
are drawn in Section 7.
2. Some outlines on grid-connected inverters
MFGCIs are special GCIs, so a brief introduce on conventional
GCIs is quite necessary [27–29]. GCIs are key components in DGSs
and MGs, and act as effective interfaces to connect distributed
RESs or micro-sources, such as photovoltaic (PV) arrays, wind
turbines (WTs), micro-gas turbines, energy storage devices and so
on, to utility grid, as shown in Fig. 1. It is worth nothing to note
that the high efficiency and low cost are two important issues of
GCIs. In general, GCIs can be classified as single-stage and
multiple-stage. Because the more stages reduce the efficiency of
a GCI much more, a multiple-stage GCI mainly has two stages.
A typical two-stage GCI is comprised of a DC/DC stage and a DC/AC
stage, as depicted in Fig. 1. The DC/DC stage is usually used to
realize maximum power point tracking (MPPT) for WT or PV
applications, or bidirectional power flow control for energy storage
application [30–32]; whereas, the DC/AC stage is used to control
the power and current injected into utility grid. Accordingly, a
single-stage GCI just has the DC/AC stage, which must complete all
the functionalities of a two-stage one had. However, a single-stage
GCI uses few electronic components, and has smaller bulk, higher
efficiency, lower cost, as well as higher reliability, compared with a
two-stage one. On the contrary, a two-stage GCI has a simpler
control algorithm since different functionalities are separated in
two independent stages. Besides, the low dc voltage of the micro-
source can be flexibly boosted by the DC/DC stage to meet the
requirement of the DC/AC stage, which is another advantage of a
two-stage GCI compared with a single-stage one.
Single-stage and two-stage GCIs have advantages and disad-
vantages of each other, so it is hard to say which ones are better.
They are all implemented in different suitable occasions. Gener-
ally, small-capacity-scale grid-connected systems are more like to
use two-stage GCIs due to their flexible feature; however, big-
capacity-scale systems mainly use single-stage GCIs for their high
efficiency and reliability. Some engineering examples are illustrated
as follows:
1. For PV application, as demonstrated in Fig. 2, the two-stage
GCIs are mainly employed in single-phase utility grid and their
capacities are usual small; however, the single-stage GCIs are
mainly utilized in three-phase utility and their capacities are
relatively bigger.
2. For WT application, the grid side converter of doubly fed
induction generator (DFIG) can be regarded as a single-stage
GCI, as indicated in Fig. 3. However, for small-capacity-scale
direct-driven permanent magnet synchronous generators
(PMSG) WT application, the two-stage GCI configuration shown
Fig. 1. Typical configuration of a grid-connected system with micro-sources.
Fig. 2. Configuration of different PV application circumstances. (a) Two-stage structure and (b) single-stage structure.
DFIG
Utility
Grid
gear
Wind Turbine
Grid-side
Converter
Machine-side
Converter
Fig. 3. Grid-connected system of DFIG.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270224

in Fig. 4(a) is a good choice as well. The DC/DC stage is fed by
the diode rectifier and tracks the maximum power point of the
WT, while the DC/AC stage connects to utility. However, for a
large-capacity-scale PMSG WT system, the single-stage GCI is
the best choice, as exhibited in Fig. 4(b).
3. For energy storage application, the single-stage and two-stage
GCIs are also suitable for different circumstances, as displayed
in Fig. 5. The energy storage cells with high enough dc voltage
can be directly fed by a single-stage PWM converter. Otherwise,
a DC/DC stage with step-up feature is necessary to match the
output voltage of cells and the input voltage of the DC/AC stage.
According to previously mentioned topologies of GCIs for RESs
application, it can be seen that either two-stage or single-stage GCI
has a DC/AC stage. The DC/AC stage is the absolutely indispensable
part of a GCI to convert the dc energy of RESs into the ac energy
and interface into utility. In this paper, as shown in the following
parts, the common DC/AC stages of the GCIs are focused, and it is
found that the DC/AC stages of GCIs can be carried out some
advanced and auxiliary functionalities to enhance the power
quality at their PCCs.
3. Power quality of distributed generation systems and micro-
grids
Due to the numerous power electronic devices, nonlinear,
unbalance and reactive local loads, the power quality at the PCCs
of DGSs and MGs maybe rather bad [33–37]. However, the power
quality of DGSs and MGs is very important issue for the stable and
economical operation of GCIs. On one hand, price of the electricity
sold to utility will be determined by its quality in a competitive
electricity market in the near future. So the power quality of DGSs
and MGs will directly relate to the price of sold electricity, and
affect their economic benefits [38–40]. On the other hand, the
power quality at PCC will seriously influence the stability of GCIs
[41–43]. Because the GCIs are mainly connected to the secondary
side of the transformers, the nonlinear loads will cause the
distortion of PCC voltage. This distorted voltage will directly
worsen the voltage and current control loops of a GCI, and lead
to distortion of its grid-connected current. In some severe cases, it
even leads to the unplanned trip off of the GCI. Besides, from the
electric power system point of view, poor power quality may result
in additional loss and overheat of power equipment, boring noises as
well torque oscillations of electric machines, the faults of sensitive
loads, and the interference of communication network [44].
Available researches on the power quality of DGSs and MGs
mainly focus on the comprehensive assessment of power quality,
advanced control strategies of GCIs in non-ideal voltage condi-
tions, and power quality management. Literature [33] has ana-
lyzed the sources of power quality problem in MGs, and has
exploited this new research field. In [35,36] the resonance phe-
nomenon in a PV plant has been studied to explain the undesired
trip off of GCIs, which shows the significant necessary of power
quality enhancement in DGSs and MGs. In comprehensive power
quality assessment field, literature [39] gives some useful
approaches to form a quantitative comprehensive indicator including
many different power quality indicators. Unfortunately, the compre-
hensive assessment can just provide a judgment of the existing
powerquality,whichmaybeusedasareferenceforDGSsandMGs
to manage their power quality. However, the bad power quality will
not be enhanced if there has no effective response to be taken. Facing
to the bad power quality of DGSs and MGs, there are two response
strategies, in general. One strategy is effective ride-through
approaches of GCIs to adapt the poor power quality [41–43].
Literature [41] has studied the operation strategy of GCIs in unba-
lance and distorted voltage conditions to enhance the quality of grid-
connected current. Obviously, this is a passive response strategy,
which cannot fundamentally change the existing bad power quality
of DGSs and MGs. Another strategy is employing active and/or
passive power quality conditioners to manage the bad power quality
ofDGSsandMGs.Amongthem,harmonicfilters and capacitors
are typical passive ones, which are good choices due to their
advantages of low cost and easy maintenance. However, the active
power quality conditioners, such as active power filter (APF) [45–48],
dynamic voltage regulator (DVR) [49–51], power factor correction
(PFC) [52,53], unified power quality conditioner (UPQC) and so on
[54,55], gain more and more applications because of their good
performance and flexibility. It is worth nothing to note that all these
power quality conditioners will cause new extra capital investment in
a DGS or MG, and need additional space, maintenance cost, and man-
hours; besides, they also may decrease the stability and reliability of
the DGS or MG.
Fortunately, active power quality conditioners have the same
essential DC/AC stage of GCIs, as shown in Fig. 6 and mentioned
before, thus these DC/AC stages can be multiplexed [47–55].
Therefore, there just a little modification in software is needed
to change the conventional GCIs into MFGCIs, in such a way that
the DC/AC stage of a GCI can be utilized to realize the function-
alities of the GCI and the power quality conditioners, as well as,
can greatly reduce the cost and bulk, and increase the cost-
effective features of the system, compared with multiple devices
with different individually independent functionalities.
4. Multi-functional grid-connected inverters in single-phase
system
MFGCI topologies in single-phase system usually have small
capacities and aim to small-scale RESs application. Available
MFGCIs in single-phase are mainly employed for PV application,
and attach APF and/or DVR functionalities.
A kind of MFGCI configuration using single-phase full-bridge
topology is given in [56–58], as illustrated in Fig. 7, whose
Fig. 4. Generation systems of direct-driven PMSG. (a) Two-stage structure and (b) single-stage structure.
Energy
Storage
System
Bi-Directional
Converter
Utility
Grid
DC/AC Stage
DC/DC Stage
Fig. 5. Grid-connected system of energy storage device.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 225

important system parameters are listed in Table 1. The whole
system consists of PV array, dc filter, dc switch SW
1
, buffer
capacitor C
dc
, single-phase full-bridge converter, filter inductor L
s
,
utility, local load, load switch SW
2
, and the controller.
This configuration can act as the interface of PV and APF at the
same time. According to the reference directions of current in
Fig. 7, either the utility or the GCI can supplies the load after load
switch SW
2
is closed. If the capacity of load is small, the surplus
power of PV, after supplied to load, will feed to utility. On the
contrary, if the capacity of load is so large that the PV can just
supply part power to it, the rest power of load will be provided by
utility. Note that if the local load is a rectifier or other nonlinear
load, the output current of the GCI can be regulated by the
controller, which can compensate the harmonic components of
load current. Therefore, the grid-connected current i
s
will be
ensured as the pure active current component. This condition is
named as real power injection (RPI) operation mode. Visibly, the
GCI also can act as an APF.
According to the topology in Fig. 7, literature [56] gives the
control strategy as indicated in Fig. 8. This MFGCI has two
operation modes, namely, maximum generation power tracking
and APF. The reference voltage of dc-link v
c,ref
is set according to
the operation mode to satisfy the requirement of APF functionality
in APF mode; however, the v
ref
is directly determined by MPPT in
maximum generation mode. Note that, G
vf
is the filter of feedback
loop. The output of dc voltage regulator G
vci
forms the amplitude
of the reference current I
m
, which multiplies the unit-quantity
synchronous signal of the utility voltage v
s
to form the instanta-
neous reference current i
s,ref
. The reference current summing the
load current i
L
yields the output current reference i
o,ref
of the
MFGCI. This reference current and the feedback current i
o
, asso-
ciated with the current regulator G
cc
, the fed-forward controller
G
vs
, the feed-forward controller G
fd
, and the compensation con-
troller G
ch
, form the modulation signal v
con
. With the help of
sinusoidal pulse width modulation (SPWM) coefficient K
PWM
, the
trigger pulses of insulated gate bipolar transistors (IGBTs) S
1
–S
4
can be obtained.
Facing to the single-phase full-bridge topology as shown in
Fig. 7, literature [57] also gives a control strategy, whose config-
uration is depicted in Fig. 9. In this strategy, there are two
operation modes as well, namely, maximum power generation
tracking and APF. In daytime, the MFGCI acts as a solar generator
and track the maximum power generation to convert solar
irradiation energy to electric energy. In day night, because the
solar irradiation is zero, the MFGCI acts as an APF to improve the
power quality at PCC. As demonstrated in Fig. 9, the control
strategy mainly consists of a self-learning unit, an inverter-
current estimator, a load current analyzer, a utility-current-
command calculator, a MPPT controller, and a dc-bus controller.
Where, the MPPT controller takes disturbance and observation
(D&O) approach [59–61], since it is simple and easy to implement
Table 1
Parameters of the typical single-phase full-bridge MFGCI.
Dc source PV array, the voltage of dc-link is V
pv
¼250 V
Capacity ≤1.5 kVA
Utility voltage 110 V/60 Hz
Switching frequency 20 kHz
Passive components L
f
¼2 mH, C
f
¼880 μF, C
dc
¼940 μF, L
s
¼5mH
Power electronic devices IGBT (Toshiba: GT15Q101, 1200 V/15 A)
Control strategy PI control, SPWM modulation
Extra functions APF, RPI
Fig. 8. Control strategy of the MFGCI configuration in [56].
Fig. 9. Control strategy of the MFGCI configuration in [57].
Utility
Grid
DC/AC Stage
Energy
Storage
System
Boost
Converter or
Bi-Directional
Converter
DC/DC Stage
Active Power Quality
Conditioner
GCI
Fig. 6. Basic components of an active power quality conditioner compared with a
GCI.
Fig. 7. Typical configuration of a single-phase full-bridge MFGCI.
Table 2
Parameters of the single-phase full-bridge MFGCI presented by Wu et al.
Dc source PV array (SOLAREX MAGA SX-60)
Capacity 1 kW
Utility voltage 110 V/60 Hz
Passive
components
L
s
¼5 mH, C
s
¼5μF, C
f
¼0.1 μF, L
r
¼1.7 mH, C
r
¼5.6 nF,
C
dc
¼470 μF, C
1
¼C
2
¼470 μF
Switching
frequency
19.45 kHz
Control
strategy
PI control, SPWM
Extra functions APF, PFC
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270226

on a digital signal processor (DSP) control board. By the means of
self-learning unit, the pseudo-linear inductance L|i
est
(t) can be
estimated, then the inverter-current i
o,est
can be formed by the
current estimator. Therefore, the current senor for inverter current
can be cancelled. Furthermore, due to the sampled utility current
i
s
, the load current i
L,est
can be obtained. Besides, the extended
instantaneous power theory is used to calculate the total power of
the load P
load
, as well as the reactive and harmonic current, i
Lq,est
and i
Lh,est
, respectively. Notation that reactive and harmonic
current limiters are employed to prevent the total reference
output current of the MFGCI exceeding its rated one, which can
generate two compensation coefficients, namely K
qc
and K
hpwc
.In
order to control the voltage of dc-bus, the reference active power
P
c
can be formed by voltage deviation regulator ΔPor MPPT
controller P
MPPT
in APF or MPPT modes, respectively. Then, the P
c
subtracting load power P
load
yields the reference grid-connected
power. And it multiplies the unit-amplitude-voltage synchronous
signal i
u
, which yields the active part of the reference current. To
achieve harmonic and reactive current compensation of the
MFGCI, it is important to analyze the estimated harmonic and
reactive current, i
Lh,est
and i
Lq,est
. It should be noted that the rated
apparent power of the MFGCI is the limited. So the total apparent
power of extra compensation power and the conventional active
power for RPI may exceed the rated apparent capacity of the
MFGCI. The simplest approach to solve this issue is to limit the
amplitude of the compensation current. In Fig. 9, coefficients 1
−K
hpwc
and 1−K
qc
are employed to yield the limited harmonic and
reactive compensation reference current. The synthetic reference
current i
s,ref
and grid-connected current i
s
are utilized in the
current regulator G
cc
to generate the trigger pulses of IGBTs
S
1
–S
4
by SPWM modulation.
Wu et al. have also described a MFGCI configuration for street
lighting application as introduced in Fig. 10, whose parameters are
described in Table 2 [62]. At daytime, the PV array is interfaced to
utility by the MFGCI. At night, the lamps attached at the dc-link
can be fed by utility and the MFGCI acts as an APF to compensate
the harmonic and reactive current. To decrease the cost, the whole
Fig. 11. Schematic control diagram of the MFGCI presented by Wu et al. (a) Control principle under grid-connected mode and (b) control principle under islanded mode.
Fig. 12. Configuration of the MFGCI investigated by Sladic et al.
Fig. 10. Configuration of the MFGCI investigated by Wu et al.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 227

system is controlled by a micro-computer unit (MCU) Intel
80C196MC.
It should be noted that the MFGCI can work in grid-connected
and islanded modes at daytime. The control schematic diagram of
the MFGCI in two different modes is given in Fig. 11(a) and (b).
Under grid-connected mode, the MFGCI supplies the solar energy
to local load firstly, and the load can absorb the rest power from
utility; of cause, the surplus power of PV can be fed to utility
as well. Therefore, the main mission of the controller in
grid-connected mode is to control the current of the MFGCI and
interface the solar energy to utility as much as possible, as shown
in Fig. 11(a). It can be found that a feed-forward controller G
rf
and a
PI controller G
cc
are employed to achieve the performance of the
system, where K
fc
is the feedback gain. On the other hand, the
MFGCI can also act as an uninterrupted power source (UPS) in
islanded mode to feed the local load when the utility is fault.
As indicated in Fig. 11(b), the controller ensures the MFGCI to be a
voltage-controlled source in such condition.
Sladic et al. have investigated a MFGCI configuration as depicted
in Fig. 12, whose parameters are presented in Table 3 [63]. Unlike a
conventional GCI, a novel variable voltage converter is embedded in
thedcsidetoadaptlargedcvoltagerangeofPVarrays[64].Besides,
an analog controller is implemented to generate the trigger pulses
of IGBTs.
Calleja and Jimenez also give a MFGCI configuration as shown
in Fig. 13, whose parameters is described in Table 4 [65].Fig. 14
demonstrates its control scheme. From Fig. 14(a) it can be seen
that three Hall Effect sensors H1, H2, and H3, are employed to
obtain the current signals for MPPT controller and power quality
conditioning. It should be noted that D&O method is utilized for
MPPT. Fig. 14(b) illustrates the block diagram of the circuit for
harmonic and reactive current detection by the means of adaptive
interfacing cancelling algorithm, where v
R
is the amplitude of the
sinusoidal reference signal. According to Fig. 14(b), the transfer
function between v
H3
and v
QD
can be expressed as
GðsÞ¼ s
5
þ2s
3
ω
2
þsω
4
s
5
þ2s
3
ω
2
þsω
4
−kω
4
ð1Þ
where k¼G
1
G
2
v
R
/τ, and τis the integral time constant.
Seo et al. also have presented a single-phase single-stage
MFGCI configuration and its control strategy, as shown in Fig. 15,
whose important parameters are available in Table 5 [66]. From
the configuration, it can be seen that, the reference current is
composed of two parts. One part I
mppt,ref
is from the result of MPPT.
Table 3
Parameters of the MFGCI presented by Sladic et al.
Dc source Voltage of PV array 150–500 V
Passive components L
1
¼8mH,L
2
¼1mH,L
3
¼1mH,C
1
¼10 μF, C
2
¼2mF
Switching frequency 15 kHz
Control strategy Hysteresis modulation
Extra functions APF
Fig. 13. MFGCI configuration presented by Calleja and Jimenez.
Table 4
Parameters of the MFGCI presented by Calleja and Jimenez.
Dc source PV array, v
pv
¼150 V
Capacity 1 kW
Passive components Transformer ratio 1:2, L
PV
¼35 mH
Switching frequency 14.2 kHz
Control strategy Hysteresis modulation, hysteresis B¼200 mA
Extra functions APF and RPI
Fig. 14. Control scheme of the MFGCI proposed by Calleja and Jimenez. (a) Schematic diagram of control strategy and (b) structure of the adaptive filter.
Table 5
Parameters of the single-phase full-bridge MFGCI proposed by Seo et al.
Dc-source PV array, the voltage of dc-bus, V
pv
¼600 V
Capacity 3 kVA
Voltage of utility grid 220 V/60 Hz
Switching frequency 20 kHz
Control strategy PI Controller, SPWM modulation
Extra function APF
Fig. 15. Single-phase full-bridge MFGCI configuration presented by Seo et al.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270228

The other part I
har,ref
is from the APF controller. Therefore, the
MFGCI can also be used as APF in day night, which can increase the
using hours of the MFGCI compared to conventional GCIs. Simul-
taneously, the trigger pulses are generated by SPWM modulation.
According to the MFGCI previously mentioned configuration,
Fig. 16 gives the detailed control structure. The phase-locked loop
(PLL) is employed to obtain the phase of utility θ
PLL
. Meanwhile,
the sampled load current is defined as I
load,a
, namely equivalent
load current of phase-ain virtual three-phase system. Then the
equivalent current of virtual phase-band ccan be yielded by 1201
and 2401phase-shift. According to the transformation from abc
frame to αβ frame, namely Clarke transformation as shown in (2)
and its inverse transformation is C
αβ=abc
¼C
−1
abc=αβ
¼C
T
abc=αβ
, the
equivalent current in αβ frame can be expressed as I
loadα
and I
loadβ
.
As a result, the load current in dq frame can be formed based on
the transformation in (3) whose inverse transformation meets
C
dq=αβ
¼C
−1
αβ=dq
¼C
T
αβ=dq
, as well the ones filtered by low pass filter
(LPF) can be written as I
lpfd
and I
lpfq
. As a consequence, the
fundamental component of the load current I
refα
can be yielded
with the aid of inverse transformations. The deviation between
I
loadα
and I
refα
is the harmonic component of the load I
har,ref
. On the
other hand, the reference voltage of the dc bus can be determined
by the MPPT controller associated with D&O approach. Due to the
dc bus voltage regulator, the maximum current amplitude of the
inverter is ensured as I
max
, which multiplies the sinusoidal wave-
form sinθ
PLL
to form the reference current I
mppt,ref
. It is worth
nothing to note that this part of reference current is the pure
active power component. To compensate the harmonic current of
the load, the detected current component I
har,ref
should be added
into I
mppt,ref
to yield the total reference current I
ref
. With the output
current regulator, the trigger pulses of the full-bridge can be
obtained by SPWM generator.
C
abc=αβ
¼ffiffiffi
2
3
r1−1=2−1=2
0ffiffiffi
3
p=2−ffiffiffi
3
p=2
"# ð2Þ
Fig. 16. Control strategy of the MFGCI configuration presented by Seo et al.
Fig. 17. The MFGCI configuration proposed by Wu et al. (a) Overall view of the MFGCI and (b) detailed block diagram.
Table 6
Parameters of the single phase full bridge topology by Wu et al.
DC-source PV array, voltage of dc-bus V
dc
¼400 V
Capacity 1 kVA
Voltage of utility grid 220 V/60 Hz
Switching frequency 25 kHz
Control strategy PI controller, SPWM modulation
Extra function APF
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 229

C
αβ=dq
¼cosθ
PLL
sinθ
PLL
−sinθ
PLL
cosθ
PLL
"# ð3Þ
Wu et al. have described a single-phase two-stage MFGCI
structure as demonstrated in Fig. 17, whose parameters are listed
in Table 6 [67]. Metal-oxide-semiconductor field-effect transistor
(MOSFET) is employed due to its fast switching frequency. It can be
found that different CPUs are employed for DC/DC and DC/AC
stages. The DC/DC stage is a typical boost circuit and a microcon-
troller PIC6F88 is implemented; on the contrary, the TMS320F2406A
DSP is used in the DC/AC stage. Note that there is a single-phase
diode rectifier connected at the output terminal of the MFGCI to act
as a nonlinear load. Meanwhile, the load is also connected at the dc-
bus of the MFGCI for reliability improvement.
There are three operation modes of this MFGCI according to the
output power of the PV array, namely grid-connected mode, direct
supply mode, and APF mode. When the solar irradiation is high, the
output power of the PV array supplies load firstly, and the surplus
power is fed to utility by the MFGCI, as depicted in Fig. 18(a).
When the solar irradiation is middle, the PV array just supplies load
and there is no ac power fed to utility grid. That is to say, the whole
system operates in direct supply mode, as shown in Fig. 18(b). As
shown in Fig. 18(c), when the solar irradiation is low, the output
power of the PV array cannot satisfy the demand of load, the shortage
power is supplied by utility and the MFGCI acts as APF to compensate
the harmonic current of nonlinear load. Table 7 shows the different
voltage levels of the boost circuit V
c
and dc-bus V
dc
in different
operation modes.
On the basis of Fig. 7, Wu et al. give a further work. As indicated
in Fig. 19, the half-bridge topology is employed to replace the full-
bridge in Fig. 7, which have the same functionalities of the before
mentioned one [68,69]. Although this half-bridge circuit can
decrease the cost of IGBTs, the voltage balance of the two series
capacitors is hard. It should be noted that, this MFGCI also has
three operation modes. When the solar isolation is high, it works
under RPI mode to convert the solar energy efficiently. When the
solar isolation is middle, it acts as a partial APF (PAPF). When the
solar isolation is low, it operates under full APF mode (FAPF).
The control strategy of the half-bridge MFGCI is shown in
Fig. 20. From this detailed block diagram, it can be seen that the
control strategy contains a MPPT controller, an inverter-current-
command calculator, a dc-bus controller, and an inverter current
controller. Some detailed outlines are similar as the control
strategy mentioned in Fig. 8. The load type and its power P
load
can be detected by load current analyzer. According to the type of
load, the MFGCI can work under APF mode or MPPT mode, while
the MPPT controller employs D&O approach. The power P
I
deter-
mined by its operation mode and the load power P
load
yield the
reference active power P
s
. The reference power multiplies the
unit-amplitude-voltage synchronous signal i
u
(t), which derives the
reference current i
s,ref
for RPI implementation. With load current i
L
feedback and inverter-current-command calculator, it is available
to generate the reference current of the MFGCI i
o,ref
. There are two
different algorithms for current regulator application to prevent
the reference current exceeding the rated one of the MFGCI,
namely amplitude clamping algorithm (ACA) and amplitude-scaling
Fig. 18. Power flow of the MFGCI configuration operates under (a) the grid-connection mode, (b) direct supply mode, and (c) APF mode.
Table 7
DC voltages of the system in different operation
modes.
Operation modes Voltage
Grid-connected mode V
c
>400 V, V
dc
¼450 V
Direct supply mode V
c
<311 V, V
dc
¼450 V
APF mode V
c
¼311 V–400 V
Fig. 19. Single-phase half-bridge MFGCI configuration presented by Wu et al.
Fig. 20. Control strategy of the MFGCI given by Wu et al.
Table 8
Controller coefficients under different operation modes.
Coefficients FAPF PAPF mode RPI mode
Linear loads Rectifier loads
ASA ACA
i
c,ref
(t)>I
sw
i
c,ref
(t)≤I
sw
K
ACA
00 010
K
ASA
1(I
sw
–I
R,pv
)/Icp 11
K
pv
11 10
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270230

algorithm (ASA). The reference current can be expressed as
i
o,ref
¼ffiffiffi
2
pðP
MPPT
þP
ACA
K
ACA
Þ
V
S,rms
i
u
ðtÞþ i
L
ðtÞ−ffiffiffi
2
pP
load
V
S,rms
i
u
ðtÞ
"#
K
PV
ð1−K
ACA
ÞK
ASA
þðI
sw
−I
R,pv
ÞK
ACA
¼i
c,p
ðtÞþi
′
c
ðtÞð4Þ
where i
c,p
(t) denotes the active component of the inverter current,
while i
′
c
ðtÞrepresents its reactive and harmonic components. i
L
(t)
is the current of the load, while P
load
is load power. P
MPPT
is the
output power of MPPT controller, and P
ACA
is the injected power using
ACA algorithm. I
sw
is the rated current of the MFGCI. I
R,pv
stands for
the active part of the output current of the PV array. It should be
noted that K
pv
,K
ACA
,andK
ASA
are controller coefficients, which are
determined by different algorithms and operation modes as illu-
strated in Table 8. The amplitude of reference current is limited by
the ACA or ASA algorithm in current regulator block, when the
reference current exceeds the rated current of the MFGCI. In
summary, the MFGCI can transfer among three different modes
seamlessly.
Patidar et al. have studied a single-phase MFGCI solution, as
demonstrated in Fig. 21, whose important parameters are shown
in Table 9 [70].Fig. 22 gives the detailed block diagram of its
control strategy. Obviously, this system is a typical current-
controlled voltage source inverter (CC-VSI) and consists of the
PV array, an H-bridge, a local load, and the filter inductor L
c
. Due to
the twice order line-frequency voltage fluctuations of dc-bus, a
large buffer capacitor is useful. Meanwhile, the filter inductor is
employed to suppress the harmonic current around the switching
frequency.
From the control strategy, a TMS320F2812 DSP is utilized as the
controller board, where the MPPT controller employs look-up table
method that saves much CPU time and storage space; never-
theless, it may deviates from the actual maximum power point in
some bad cases. Because of the sampled voltage and current of PV
array, I
pv
and I
pv
, and the calculated maximum output power of PV
P
pv
, the reference voltage of PV array V
dc,ref
can be obtained by
the MPPT controller. Due to the dc-bus voltage controller, the
Fig. 21. Single-phase MFGCI configuration presented by Patidar et al.
Table 9
Parameters of the single phase full bridge topology by Patidar et al.
Dc-source 15 series and 1 parallel, BPSOLAR BP280 PV array
Capacity Approximate 1.2 kVA
Voltage of utility grid 230 V/60 Hz
Switching frequency 25 kHz
Passive components L
c
¼3mH,R
c
¼0.1 Ω,L
L
¼25 mH, R
L
¼7.5 Ω
Control strategy PI control, hysteresis modulation
Extra function APF
Fig. 22. Control strategy of the single-phase MFGCI proposed by Patidar et al.
Fig. 23. Brief topology and control strategy of the MFGCI presented by Hirachi, etc. (a) Overview of the MFGCI and (b) control strategy of the MFGCI.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 231

reference power P
sl
is achieved, meanwhile the average active
power of load P
L
subtracting P
sl
derives the grid-connected
reference power. With the help of optimal control, it is easy to
form the amplitude of the grid-connected current I
sm
¼P
s
/V
sm
,
and it multiplies the unit-amplitude-voltage synchronous signal to
yield the instantaneous reference current i
s,ref
. Then, the trigger pulses
of the IGBTs are achieved by the grid-connected current feedback and
current regulator associated with hysteresis modulation.
According to the topologies mentioned before, these MFGCI
using full-bridge or half-bridge in single-phase have three visible
drawbacks:

Firstly, due to the large fluctuations of solar irradiation, these
MFGCI topologies can solely operate under grid-connected
mode. If they work under islanded mode and the output power
of the PV array is larger than the rated power of the local load,
the terminal voltage of the load will be bigger than the rated
one. On the contrary, when the output power of the PV array is
smaller than the local load, the terminal voltage of load will be
smaller than the rated one. In summary, the system is hard to
stably work at its nominal point in islanded mode.

Secondly, the previously mentioned MFGCIs mainly employ
L-filter, which need a big inductor and the performance is bad.
Furthermore, the cost and bulk of the filter inductor will
increase if the inductance is bigger.

At last, there is no electric isolation, the dc component injects
into utility will affect other devices. Especially, the dc compo-
nent may lead to the saturation of transformer due to the dc
bias magnetic.
The single-stage single-phase MFGCI topology presented by
Hirachi et al. in Fig. 23, can overcome the drawbacks above, whose
parameters are listed in Table 10 [71]. In this topology, the diodes
D
1
and D
2
in Fig. 7 can be cancelled because of the energy storage
device embedded in the dc-link side. In addition, a LC-filter is
implemented to replace the L-filter, and an isolation transformer is
attached in the ac side. It should be noted that there are three
operation modes. In sunny days, the bi-directional converter feeds
the energy of PV array to local load and utility, as shown in Fig. 24
(a). In cloudy day or the interruption of utility, the system work
under islanded mode, and the battery supplies power to load, as
shown in Fig. 24(b). At day night, the converter acts as a PWM
rectifier and charges the battery, as shown in Fig. 24(c).
From Fig. 24, it can be found that the system can stably work
under islanded mode. It can be seen that the system is a simple
micro-grid. When the grid switch is close, the whole system works
under grid-connected mode. The output power of PV array
supplies the local load firstly. Then, the surplus PV power feeds
to the battery and utility; on the contrary, the shortage power is
supplied by utility and the battery. When the grid switch is open,
the system works under islanded mode. The battery can absorbs
(or supplies) the surplus (or shortage) power of the PV array,
which can keep the terminal voltage of the load at its nominal
level. As mentioned before, the system can also be modified as an
APF to feed harmonic current to local nonlinear load.
As demonstrated in Fig. 25, Dasgupta et al. have investigated a
MFGCI configuration for harmonic and reactive compensation in a
micro-grid, whose parameters are illustrated in Table 11 [72].
In this MFGCI, a CC-VSI is implemented as the interface. It should
be noted that the battery storage is attached in the dc link, so the
system can work under islanded condition freely.
Fig. 26 depicts the control scheme of the MFGCI. It can be seen
that the Hilbert transform and extended p–qpower theory are
utilized to detect the compensation current components. Besides,
a Lyapunov-based controller and spatial repetitive controller are
embodied for excellent dynamic and steady performances on
current tracking.
Chiang et al. have studied a MFGCI as shown in Fig. 27, whose
parameters are described in Table 12 [73]. From the control part of
the configuration, it can be found that a novel MPPT algorithm is
utilized to attain high performance, which has considered the
dynamic model of the PV array and the state-averaged mode of the
DC/DC converter. Besides, the charging and discharging control of
the batteries are clearly separated, as well as these modes can be
Table 10
Parameters of the single-phase full-bridge MFGCI studied by Hirachi et al.
Dc-source PV array, battery, voltage of dc-bus is 200 V
Capacity 3 kVA
Voltage of utility grid 110 V/60 Hz
Switching frequency 40 kHz
Power electronic device MOSFET
Control strategy PI control, SPWM modulation
Extra function UPS
Fig. 24. Three operation models of solar photovoltaic power generation conditioner with bidirectional power converter. (a) On clear sky day (utility interactive operation),
(b) on cloudy sky day or power failure (stand alone operation) and (c) during late-night (rectifier operation).
Fig. 25. Configuration of the MFGCI presented by Dasgupta et al.
Table 11
Parameters of the MFGCI presented by Dasgupta et al.
Dc source V
dc
¼100 V
Utility voltage 50 V
Sampling frequency 10 kHz
Control strategy Lyapunov based and spatial
repetitive control, SPWM modulation
Extra functions APF
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270232

easily transferred with the help of the operation modes control
section.
Bojoi et al. also survey a MFGCI configuration as illustrated in
Fig. 28, whose important parameters are presented in Table 13
[74,75]. Additionally, its control strategy is given in Fig. 29.
Fig. 29 shows the detailed control diagram of the MFGCI
configuration presented by Bojoi et al. It can be seen that this
control strategy consists of reference current generator, current
regulator, and SPWM modulation, where a filter algorithm asso-
ciated with the instantaneous power theory, named as sinusoidal
signal integrator (SSI), is employed to detect the harmonic and
reactive current of the load. Furthermore, a repetitive controller
and a PI controller are implemented to accurately track the
reference current.
Cirrincione et al. have also investigated a MFGCI configuration
as shown in Fig. 30 and Table 14 [76]. Simultaneously, Fig. 31 gives
its control principle with two neural adaptive filters to extract the
compensation current of local loads and compute the fundamental
component of the utility voltage. Besides, a proportional-resonant
(PR) controller is utilized to achieve excellent performance.
Macken et al. have studied the single-phase single-stage MFGCI
system as displayed in Fig. 32, whose parameters are listed in
Table 15. Whereas, there may have multiple MFGCIs in a DGS or
MG, thus they have presented an agent-based communication
approach to coordinate the MFGCIs as described in [77]. In general,
to prevent the output current of the MFGCI exceeding its rated
one, its apparent power for auxiliary services cannot be too much.
Fortunately, a DGS or MG always has integrated many GCIs. If they
can work as MFGCIs coordinately, the total apparent power may be
large enough to deal with the possible power quality issues.
Fig. 26. Control strategy of the MFGCI proposed by Dasgupta et al. (a) Block diagram in detail and (b) the subsystem of the control scheme to generate reference current.
Fig. 27. Architecture of the MFGCI investigated by Chiang et al.
Table 12
Parameters of the single-phase full-bridge MFGCI presented by Chiang et al.
Dc source PV array, 7 6 (each with open voltage
V
oc
¼14 V, short current I
sc
¼1.2 A )
batteries 12 V/15 Ah14
Capacity Approximately 60 0 W
Utility voltage 110 V/60 Hz
Passive components L
p
¼0.3 mH, C
i
¼200 μF
Power electronic components MOSFET 450 V/13 A (boost converter),
IGBT 60 0 V/50 A (inverter)
Control strategy PI control, SPWM
Extra functions APF, UPS
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 233

The MFGCI configurations mentioned before are all used as
power quality conditioners in parallel, which can just enhance the
power quality issues caused by harmonic and reactive current.
Hosseini et al. present a novel single-stage MFGCI topology in
series, which can act as a DVR and compensate the voltage swell
and sag of utility [78]. Important parameters of the MFGCI
configuration are described in Table 16.
As shown in Fig. 33, this topology is made up of two indepen-
dent boost circuits. The desired voltage v
o
can be achieved at its
output terminal if a proper control strategy is carried out. When
the voltage of utility is swell, sag, and/or interruption, this
topology can compensate it effectively. In addition, this topology
can also compensate the reactive voltage at the terminal of load,
which can confirms unit factor operation of utility. It should be
noted that the filter components L
dc
,L
f
, and C
f
pay important role
to suppress the fluctuations of the dc-link and inhibit the failure of
Fig. 28. The MFGCI configuration presented by Bojoi et al.
Table 13
Parameters of the single-phase full-bridge topology by Bojoi et al.
Dc-source Micro-source, V
dc
¼400 V
Capacity 4 kVA
Voltage of
utility grid
220 V/50 Hz
Switching
frequency
10 kHz
Passive
components
L
F
¼0.7 mH, C
F
¼7.5 μH, C
F
¼2.2 mF, R
2
¼25 Ω,L
2
¼30 mH,
L
1
¼0.9 mH, C
1
¼1mF,R
1
¼50 Ω
Control strategy Repetitive control and PI control, SPWM modulation
Extra functions APF, reactive compensation
∫
∫
Fig. 29. Block diagram of the MFGCI proposed by Bojoi et al. (a) Overview block diagram, (b) reference current generation algorithm, (c) model of the SSI filter and
(d) current regulator.
Fig. 31. Control diagram of the MFGCI presented by Cirrincione et al.
Table 14
Parameters of the single-phase full-bridge MFGCI presented by Cirrincione et al.
Dc source PV array, V
dc
¼250 V
Utility voltage 130 V/50 Hz
Passive components L
g
¼20 mH, R
g
¼1Ω,L¼4mH,R¼0.2 Ω
Switching frequency 15 kHz
Control strategy PR control, SPWM
Extra functions APF
Fig. 30. Schematic diagram of the MFGCI investigated by Cirrincione et al.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270234

the MPPT. In the presented topology, each boost circuit is driven
by a modulation signal with 1801phase-shift and dc bias.
The deviation of these two modulation signals yields a sinusoidal
waveform. Therefore, the terminal voltage of this topology can
be controlled by the modulation signals directly. As shown in
Fig. 33(a) and (b), the controller of the boost circuit consists of an
outer current loop, and an inner voltage loop, respectively. If the
desired output voltage of this topology can be expressed as
v
oref
¼ffiffiffi
2
pVsinðωtÞ, then the reference voltage of two boost circuits
can be written as
(1) when sin(ωt)>0, take
v
o1ref
¼v
o2
þv
oref
¼v
o2
þffiffiffi
2
pVsinðωtÞ
v
o2ref
¼v
dc
−VsinðωtÞ=ffiffiffi
2
p
8
<
:ð5Þ
(2) when sin(ωt)<0, set
v
o1ref
¼v
dc
þVsinðωtÞ=ffiffiffi
2
p
v
o2ref
¼v
o1
−v
oref
¼v
o2
−ffiffiffi
2
pVsinðωtÞ
8
<
:ð6Þ
Fig. 34 shows the control block of the boost circuit, and Fig. 34(a)
give the algorithm to generate the reference voltage. The reference
voltage of dc-bus V
pref
is achieved by MPPT controller. And
the reference maximum output power P
MMP
is obtained by dc
voltage regulator. In addition, the reference phase φ
0
can be
accurately calculated. To improve the dynamic performance on
power tracking, a close-loop control of P
INV
is added for the phase
compensation. Then, the amplitude of the reference voltage |v
L
|can
be derived according to different operation modes. Furthermore, by
the means of the outer voltage loop and inner current loop, as
shown in Fig. 34(b) and (c), the controlled duty cycle of the boost
circuits can be achieved.
Because the impedance of utility grid is small, to compensate
utility voltage using a parallel converter is unwise, in general.
Otherwise, there may be a very large current flow across utility,
that is to say, the capacity of the parallel converter must be very
large. However, in the view point of the PV GCI, this kind of
solution may be suitable, because the capacity of the PV inverter is
usually large enough to supply the full capacity of local load.
Therefore, a properly large reactive current can change the voltage
of load terminal, if a large inductor L
s
is introduced in utility side.
Mastromauro et al. present this kind of MFGCI as demonstrated in
Fig. 35 [79,80], whose parameters are listed in Table 17. It should
be noted that the harmonic current is so large, and can cause the
distortion voltage on utility inductor. Fortunately, it also can be
compensated by the PV GCI.
The control block of this MFGCI is depicted in Fig. 36(a), which
consists of an outer voltage loop and an inner current loop.
Furthermore, the outer voltage loop employs repetitive controller
in Fig. 36(b), which can accurately tracks fundamental and
harmonic voltage and effectively compensates the voltage swell,
sag, and distortion. Besides, the inner current loop implements PI
controller, which has a fast dynamic performance.
According to the MFGCI systems in Fig. 35, literature [81] gives
a reference generation algorithm based on droop control, which is
indicated in Fig. 37. It can be seen that the reference voltage is
formed by the reactive and active power droop loops, where the
normal values of reactive and active power are Q
n
¼0 and
P
n
¼P
MPPT
, respectively, and Q
G
and P
c
are the actually output
reactive and active power of the inverter. The detailed block
diagram of the droop control can be found in Fig. 38, where m
i
,
m
p
, and n
p
are control coefficients. The decoupling strategy of
active and reactive power can be expressed as
P′¼ðP
c
−P
n
Þsinθ−ðQ
G
−Q
n
Þcosθ
Q′¼ðP
c
−P
n
ÞcosθþðQ
G
−Q
n
Þsinθ
(ð7Þ
Fig. 32. Configuration of the MFGCI studied by Macken et al.
Table 16
Parameters of the single-phase MFGCI proposed by Hosseini et al.
Dc-source PV array
Capacity P
L
¼2–2.4 kW(R
L
¼20–25 Ω)
Voltage of utility grid 220 V/50 Hz
Switching frequency 20 kHz
Passive components L
1
¼L
2
¼0.2 mH, C
f
¼1000μF, C
1
¼C
2
¼20 μF,
L
f
¼2.5 mH, L
dc
¼1mH
Control strategy PI control, SPWM modulation
Extra functions DVR and PFC
Table 15
Parameters of the single-phase full-bridge MFGCI presented by Macken et al.
Dc source PV array
Capacity 1 kW
Utility voltage 110 V/50 Hz
Passive
components
L
1
¼5mH,C
f
¼4μF, L
f
¼2.5 mH; diode rectifier L¼150 mH,
R¼35 Ω
Switching
frequency
10 kHz
Control strategy PI control, SPWM
Extra functions APF
Fig. 33. Single-phase MFGCI presented by Hosseini et al. (a) Schematic configuration and (b) circuit model.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 235

where θ¼X/Ris impedance ratio. It should be noted that the
stability of the droop controller is directly determined by the
control coefficients. It is very important to properly design these
coefficients.
Based on the MFGCI presented by Dasgupta et al. as illustrated
in Fig. 25, they also give another MFGCI configuration as shown in
Fig. 39, whose parameters are listed in Table 18 [82]. It is easy to
found that the load is connected between PWM inverter and
utility in series. Thus, the voltage satisfy v
inv
¼v
L
−v
g
, in other
words, the voltage of load v
L
can be controlled by the MFGCI
indirectly. Therefore, the MFGCI can greatly enhance the voltage
quality of the load.
Fig. 40 shows the control scheme of the MFGCI. According to
the phasor diagram in Fig. 39(b), the reference voltage of the load
can be calculated as shown in Fig. 40. Then, a spatial controller is
utilized to obtain the trigger pulses of the IGBTs.
Lin et al. have investigated a MFGCI as shown in Fig. 41, whose
parameters are listed in Table 19. On the basis of the single-phase
H-bridge converter, an asymmetrical leg is employed to achieve
three-level PWM, which has less voltage harmonic generated on
the ac side of the MFGCI compared with a two-level one.
An independent cell of this three-level converter is described in
Fig. 41(a). According to the cell and its control scheme, it can be
applied to three-phase utility as three independent VSIs as shown
in Fig. 41(c). However, it can also be applied as a combined
inverter as shown in Fig. 41(d), whose three cells share the dc
bus fed by energy storage devices and/or RESs. In [83], a single-
phase MFGCI is studied, which can act as an UPQC, as shown in
Fig. 42.
Geibel et al. also give a single-phase UPQC-based MFGCI as
presented in Fig. 43 and have achieved good performance [84].
Kuo et al. have investigated a MFGCI configuration as shown in
Fig. 44, whose system parameters are listed in Table 20.In[85], the
linear relationship between equivalent conductance and current of
PV array is found. As a consequence, a novel MPPT algorithm is
presented so as to perform the rapid and accurate power tracking
features. Owing to the 3-leg topology, the MFGCI can be applied to
single-phase three-wire system.
The control schematic diagram of the MFGCI is depicted in
Fig. 44. From the control principle of the line-mode controller in
Fig. 45(a), it can be seen that the MFGCI can performs as a
conventional GCI or an APF flexibly. The amplitude of the reference
current is derived by MPPT controller or APF controller, which
multiplies the per-unit signal in phase with utility voltage to
generate the instantaneous reference current i
ulm,ref
. It should be
noted that the subscript “lm”denotes the line-value, for instance
i
lm
¼i
a
−i
b
. Then, the load current i
Llm
is added to i
ulm,ref
for output
current regulation of the MFGCI. Besides, a phase-lead controller
G
cc
, a feed-forward controller G
fd
, and a feedback compensator G
ch
are employed to achieve fast dynamic and accurate steady perfor-
mance. To against the disturbance from utility voltage, a distur-
bance immunizing controller G
vs
is utilized. To balance the
currents in phase-aand b, a neural-mode controller is implemen-
ted as shown in Fig. 45(b).
All the MFGCI topologies mentioned before employ hard-
switching, which may lead to low efficiency of energy conversion.
De Souza et al. give a MFGCI topology using soft-switching technol-
ogy, whose configuration and parameters are demonstrated in Fig. 46
and Table 21,respectively[86–88]. A brief diagram of the topology is
Fig. 34. Control strategies of the boost circuits. (a) Inverter voltage reference controller, (b) voltage control loop and (c) current control loop.
Fig. 35. Block-diagram of the grid-connected PV system with active filter functionality.
Table 17
Parameters of the three phase full bridge topology by Mastromauro et al.
Dc-source PV array, voltage of dc-bus, V
dc
¼460 V
Capacity Danfoss VLT 5006 inverter, rated apparent power 7.6 kVA,
experimental capacity 1.2 kW
Voltage of
utility grid
220 V/50 Hz
Switching
frequency
20 kHz
Parameters LC filter 1.4 mH+5μF, damping resistance R¼1Ω, inductor in
utility grid side L
s
¼15 mH
Control strategy Repetitive control, PI control, SPWM modulation
Extra functions DVR, harmonic voltage compensation
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270236

shown in Fig. 47. It can be found that this topology consists of a half-
bridge-zero-voltage DC/DC converter and a cascaded DC/AC conver-
ter. Meanwhile, to reduce the electromagnetic capacity and the
power loss of the system, the DC/DC stage employs half-bridge
zero-voltage switching pulse width modulation (HB ZVS-PWM), as
illustrated in Fig. 48.Literature[86–88] show the algorithm to select
the optimal parameters of capacitor C
e1
/C
e2
, inductor L
r
,andthe
transformer turns ratio.
Fig. 49 gives the block diagram of the DC/AC stage. Similar to
the single-phase full-bridge MFGCIs mentioned before, this DC/AC
stage can also be implemented as an APF. This topology can not
only be an interface to connect PV array to utility, but also can act
as an APF to compensate the harmonic and reactive current of the
local load. Its operation can be briefly described as follows.
The DC/DC converter is employed to boost the output voltage of
PV array to be a high enough one to meet the requirement of the
DC/AC stage. At the same time, the DC/DC stage also completes the
MPPT of PV array. The PV array supplies load firstly, and the
surplus power is fed to utility in sunny days; on the contrary, the
utility will supply the shortage power of load, when PV array
generates not enough power to load in cloudy days. Note that, to
ensure the unit power factor of the utility, the MFGCI also can
operate as an APF to compensate the harmonic and reactive power
of the load. The control strategy of the DC/AC stage mainly consists
of an outer dc-bus voltage loop and an inner current loop. The
amplitude of grid-connected current is derived by the reference
and feedback dc-bus voltage, V
ref
and k
v
V
dc
, as well as the voltage
regulator C
v
. The amplitude multiplying the unit-amplitude-
voltage synchronous signal yields the reference current. With
the help of current regulator C
i
and forward back of utility voltage
G
cd
, the trigger pulses of the full-bridge can be obtained by SPWM
modulation.
5. Multi-functional grid-connected inverters in three-phase
system
As mentioned before, the capacities of the single-phase MFGCIs
are low, which are mainly used in residential PV systems.
In addition, the harmonic detection approaches of the single-phase
MFGCIs are harder than three-phase MFGCIs. It is worth nothing to
note that a single-phase MFGCI is a typical unbalance source, which
will burden the utility to manage the unbalance issue. Therefore,
three-phase MFGCIs have many good performances and are suitable
for general application. The available MFGCIs in three-phase system
arealsosomesingle-stageortwo-stageGCIsassociatedwithAPF,
PFC, DVR and/or UPQC functionalities.
To broaden the application field of the MFGCIs, Wu et al. give a
three-phase MFGCI, as shown in Fig. 50, based on the single-phase
one in Fig. 7, whose parameters are described in Table 22 [89].Asa
consequence, a control strategy is also presented as depicted in
Fig. 51. This configuration can accomplish the interface function-
ality of conventional GCI and the functionality of harmonic and
reactive compensation. There are three different operation modes
according to the solar irradiation, namely FAPF mode, PAPF mode,
and RPI mode. In low irradiation, the MFGCI works under FAPF
mode, and there is little active power generated by the PV array.
As a result, the MFGCI has enough apparent power to carry out the
APF functionality. When the irradiation is middle, the MFGCI
works under PAPF mode, namely, it can just supply partial reactive
and harmonic current to compensate the local load. When the
irradiation is high, the MFGCI works in RPI mode and just generate
active power. Literature [89] gives the control strategy as shown in
Fig. 52, in which, the detailed harmonic and reactive current
detection approach is given as shown in Fig. 53.
As shown in Fig. 52, a three-phase system can be viewed as two
single-phase systems. With the aid of the load current analyzer,
the average active and reactive power of the two load ports, p
L1
,
q
L1
,p
L2
, and q
L2
, as well as the reactive and harmonic current, i
Lj,h
and i
Lj,q
, can be achieved. To prevent the reference current
exceeding the rated one of the MFGCI, a limiter is employed to
achieve the reactive and harmonic coefficients, k
qc1
(k
qc2
) and
k
hpwc1
(k
hpwc2
). Then the harmonic and reactive reference current
of each phase, i
cj,h,ref
and i
cj,q,ref
, can be calculated. Similarly, the
active reference current i
cj,p,ref
is available by the means of MPPT
and voltage regulator of dc-bus. These three parts constitute the
reference current i
cj,ref
. Due to the feedback of output current and
the current regulator G
cc
, the trigger pulses of the three-phase
H-bridge can be generated by SPWM modulation.
Fig. 36. Full-bridge MFGCI configuration presented by Mastromauro et al. (a) Voltage control of the shunt connected PV converter and (b) repetitive control of voltage loop.
Fig. 37. Block diagram of the grid-connected PV-system power stage and its
control scheme.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 237

Literature [90] founds that the reference current of MFGCI can
be simplified as
i
cu,ref
¼ðP
MPPT
−P
L
ÞuðtÞ=½ffiffiffi
2
pV
rmsðuvÞ
þi
Lu
i
cv,ref
¼i
Lv
i
cw,ref
¼−ðP
MPPT
−P
L
ÞuðtÞ=½ffiffiffi
2
pV
rmsðuvÞ
þi
Lw
8
>
>
<
>
>
:
ð8Þ
where P
L
¼p
L1
þp
L2
is active power of load. V
rms(uv)
denotes the
root-mean-square (RMS) value of line-voltage. u(t) Represents the
unit-amplitude-voltage synchronous signal. Furthermore, an algo-
rithm named as fast-zero-phase detection is proposed to generate
the unit synchronous signal u(t) to form the reference current.
He et al. also have researched a three-phase H-bridge MFGCI
system as illustrated in Fig. 54, whose parameters are listed in
Table 23 [91,92]. As known to all, the previously mentioned
MFGCIs are mainly CC-VSI as shown in Fig. 54(a) typically. How-
ever, if a MFGCI can behave as a voltage-controlled VSI, many good
performances can be achieved, such as seamless transfer from
grid-connected mode to islanded mode, plug-and-play, supporting
the voltage and frequency of the DGSs and/or MGs using droop
controller, and power sharing of MFGCIs in islanded mode. There-
fore, He et al. give a voltage control scheme for the MFGCI as
shown in Fig. 54(b). It should be noted that the sliding discrete
Fourier transform (SDFT) is utilized both in voltage control mode
or current control mode to detect the harmonic voltage or current
for compensation.
Yu et al. also give a MFGCI configuration as depicted in Fig. 55
and Table 24 [93]. It can be found that the derivation regulating of
dc voltage derives the reference current in d-axis, as well as the
instantaneous reactive theory is implemented to detect the
reactive and harmonic current for compensation. The MFGCI can
realize active power generation and reactive current compensa-
tion simultaneously at daytime. Meanwhile, it can still behave as
an APF at night to enhance the power quality of utility.
As previously mentioned, the output power of the PV array and
other RESs is random and uncertain. So the whole system may be
instable in islanded mode if there is no energy storage device in
the MFGCI.
Kim et al. have given a new MFGCI configuration as demon-
strated in Fig. 56, whose parameters are given in Table 25 [94].
There is a DC/DC stage adding to the dc-bus of the MFGCI, which
can realize MPPT feature. Additionally, due to the extra battery
device at dc-bus, the system can stably operate in islanded mode,
which can be viewed as an UPS. Besides, the LC-filter brings better
performance to suppress switching harmonic current compared
with an L-filter.
Fig. 38. Block diagram of reference generation algorithm.
Fig. 39. Configuration of the MFGCI investigated by Dasgupta et al. (a) Schematic diagram and (b) detailed explanation on the phasors diagram of voltages.
Table 18
Parameters of the MFGCI presented by Dasgupta et al.
Dc source PV array, V
dc
¼270 V
Utility voltage 110 V/50 Hz
Sampling frequency 10 kHz
Passive components R¼150 Ω,L¼0.1 H
Control strategy Spatial repetitive control, SPWM modulation
Extra functions Harmonic voltage compensation, DVR
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270238

When the grid-connected switch is closed, the PV array
supplies the local nonlinear load. The utility will absorbs (or
supplies) the surplus (or shortage) active current. On the other
hand, when the switch is open, the battery will supply the local
load and the MFGCI can act as an UPS.
Fig. 57 shows the control strategy of the MFGCI configuration
proposed by Kim et al.. The instantaneous reactive power theory is
used to detect the compensation current of load. The sampled
three-phase voltage and current can be utilized to calculate the
instantaneous power of utility. Then their average active power
can be obtained by the means of LPF, which multiplies a coefficient
kto derive the amplitude of fundamental active component of load
current. The amplitude times the unit-amplitude-voltage synchro-
nous signal, and then it derives the active part of load current.
The load current subtracting the detected fundamental one, yields
the harmonic and reactive current of load I
C
. On the other hand,
with the help of MPPT and voltage regulator of dc-bus, it can yield
Fig. 40. Control scheme of the MFGCI presented by Dasgupta et al.
Fig. 41. Topology and its control scheme of the MFGCI cell presented by Lin et al. (a) Topology of the cell, (b) control scheme of the cell, (c) application in three-phase utility:
Case 1 and (d) application in three-phase utility: Case 2.
Table 19
Parameters of the MFGCI presented by Lin et al.
Dc source PV array, the voltage of dc-link is V
dc
¼200 V
Capacity 1.5 kW
Utility voltage 110 V/60 Hz
Passive components L¼2mH, C
1
¼C
2
¼110 0 μF, v
C1
¼v
C2
¼100 V
Switching frequency 20 kHz
Control strategy PI control, SPWM modulation
Extra functions UPQC
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 239

the amplitude of grid-connected current I
ref
. To control the output
current of the MFGCI, the current feedback I
in
and SPWM modula-
tion are employed, too.
Based on their previous work in single-phase system, Dasgupta
et al. have researched a MFGCI configuration in three-phase
system as shown in Fig. 58, whose parameters are described in
Table 26. The control strategy of the MFGCI is illustrated in Fig. 59.
To detect the compensation current component, a novel algorithm
is proposed in [95,96] as well.
As mentioned before, to suppress the dc bias of GCI that may be
inject into utility, the isolation transformer in ac side is necessary
sometimes. Cheng et al. give a three-phase MFGCI configuration,
as demonstrated in Fig. 60, whose parameters are shown in
Table 27 [97]. It can be seen from Fig. 60 that the load and the
MFGCI are isolated by isolation transformers in ac side, which will
increase the cost and bulk of the system. Compared with Fig. 56,it
can also be found that the energy storage device in dc side is
cancelled.
Fig. 61 gives the algorithm to calculate the reference current of
the MFGCI. It is obvious that the instantaneous power theory is
employed to detect the harmonic components of load current.
The utility voltage and current in stationary αβ frame, u
αβ
and i
αβ
,
are obtained by the means of sampled PCC voltage and load
current. With the help of instantaneous power theory, the funda-
mental component of load current i
αβf
can be achieved. As a result,
the fundamental component of load current in natural abc frame
i
babc
can be formed after Clarke transformation. The load current
i
Labc
subtracts i
babc
, and the result yields the harmonic current i
habc
.
On the other hand, according to the reference power generation of
the MFGCI, Pand Q, it can generate the normal grid-connected
reference current i
gabc
by the means of (9). The detected harmonic
current and the computed grid-connected reference current con-
stitute the total reference current of the MFGCI, as shown in
Fig. 60. Note that the transformations Tand C
pq
in Fig. 61 are
(9) and (10), respectively.
i
αref
¼ðu
α
Pþu
β
QÞ=ðu
2
α
þu
2
β
Þ
i
βref
¼ðu
α
P−u
β
QÞ=ðu
2
α
þu
2
β
Þ
8
<
:ð9Þ
p¼u
α
i
α
þu
β
i
β
q¼u
β
i
α
−u
α
i
β
(ð10Þ
Naderi et al. have studied a MFGCI configuration as illustrated
in Fig. 62, whose parameters are listed in Table 28 [98]. From the
single-line diagram, it can be found that the step-up transformer is
used to reduce the voltage stress of the converter. Besides, the
control principle mainly consists of two parts. One is the reference
generator, and another is the core controller, which is shown in
Fig. 63. From the block diagram of the controller, it is easy to see
that the instantaneous reactive power theory is utilized to com-
pute the compensating current components. Besides, hysteresis
modulation is implemented so as to achieve fast dynamic perfor-
mance of the MFGCI.
Mohod et al. have investigated a MFGCI to connect micro-wind
generator to utility and compensate the harmonic current of
local load, whose parameters are described in Table 29 [99].
The configuration of the MFGCI is shown in Fig. 64, as well as its
control strategy is given in Fig. 65. From Fig. 64, it can be seen that
the dc-link of the MFGCI is fed by wind generator and storage
battery. As a result, the battery can substitute the wind generator
and supply the local critical load uninterruptedly when the micro-
wind generator is cut off. Specially, the MFGCI can compensate the
harmonic current of the local load as well. Therefore, the power
quality at the PCC can be greatly enhanced.
To achieve the performances of the MFGCI, its control scheme
is demonstrated in Fig. 65. From Fig. 65, it is easy to find that a
Fig. 43. Configuration of the MFGCI studied by Geibel et al.
Fig. 44. Configuration of the MFGCI presented by Kuo et al.
Fig. 42. Single-phase UPQC like MFGCI presented by Lin et al.
Table 20
Parameters of the single-phase full-bridge MFGCI presented by Kuo et al.
Dc source PV array, V
PV
¼238 V
Capacity 1 kW
Utility voltage 110 V/60 Hz
Passive
components
Load: Z
1
rectifier load 100 W, Z
2
resistive load 100 W, Z
3
resistive load 100 W; input filter capacitor C
i
¼470 μF; LC
series filter L
f
¼1.6 6 mH , C
f
¼1000μF; output filter
L¼1.2 m H
Power electronic
devices
IGBT HG20N60
Switching
frequency
18 kHz
Control strategy PI control, SPWM
Extra functions APF
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270240

DFIG is implemented to catch the wind energy, and a three-phase
H-bridge topology is employed to be the energy interface. With
the help of dc link voltage PI controller, the amplitude of the
reference current iis obtained. The amplitude multiplies the unity-
quantity signal u
sabc
to derive the reference current to inject into
utility. By the means of the PI controller and the calculator for
compensation current, the trigger pulses of IGBTs can be achieved
by hysteresis modulation.
Marei et al. also give a three-phase H-bridge MFGCI configura-
tion, and present a harmonic detection approach to generate its
reference current named as multi output adaptive linear (MO-
ADALINE), as displayed in Fig. 66 and Table 30 [100,101].
The active and reactive power generation of the MFGCI is con-
trolled by i
d
and i
q
, respectively. When the MFGCI works under
FAPF mode, a fuzzy logic controller (FLC) is employed to control
the voltage of dc-bus and enhance the robustness of the system.
Meanwhile, MO-ADALINE approach is implemented to detect the
compensation component of the load current I
comp
as shown in
Fig. 67. It can be found that, the detected harmonic is feedback to
modify the linear weights of neurons. When the algorithm is
convergent, the outputs of neurons would be the harmonic current
of load exactly. It should be noted that a limiter is carried out to
prevent the reference current exceeding the rated one of
the MFGCI.
Cheng et al. also depict a MFGCI configuration, as presented in
Fig. 68, whose parameters are shown in Table 31 [102]. It can
compensate the unbalance current of local load, using a droop
control strategy, as shown in Fig. 69. It can be found that the
control strategy employs “positive-sequence active power and
frequency droop”, and “positive-sequence reactive power and
voltage amplitude droop”to form reference voltage. In addition,
a“negative-sequence reactive power and conductance droop”is
utilized to form the negative-sequence reference current. To accom-
plish these droop controllers, Park transformations in positive- and
negative-sequence frames are applied to voltage and current.
Lv et al. have given a MFGCI configuration as demonstrated in
Fig. 70, whose parameters are described in Table 32 [103,104].
From the control strategy in Fig. 71, it can be seen that i
d
–i
q
approach is implemented to detect the compensation current, and
the dc-link voltage is control by the means of d-axis current i
d
.
To mitigate the unbalance voltage in a DGS using GCI, Mohamed
and El Saadany have surveyed a MFGCI system as illustrated in Fig. 72,
Fig. 46. Single-phase MFGCI with two stages for PV application presented by De Souza et al.
Table 21
Parameters of the single-phase two-stage MFGCI presented by De Souza et al.
Dc-source PV array
Capacity 1 kVA
Voltage of utility
grid
220 V/60 Hz
Passive
components
Number of PV array is 20, rated voltage and current of each
PV array are 83.5 V and 12 A, respectively, whose short-
current (I
s
) and open-voltage (V
oc
) are 12.4 A and 107 V,
respectively.
Output voltage of DC/DC converter is 400 V, whose
switching frequency is 100 kHz. C
in
¼C
fHF
¼1000μF,
C
e1
¼10 μF, C
e2
¼5μF, L
fHF
¼50 μH, L
r
¼680 nH, C
o
¼1000μF
Switching frequency of DC/AC stage is 20 kHz,
C
fLF
¼100 0 μF, L
fLF
¼1.6 m H, L
o1
¼L
o2
¼1.2 m H
Power electronic
devices
MOSFET and IGBT
Control strategy SPWM modulation
Extra functions APF
Fig. 47. Schematic configuration of the MFGCI presented by De Souza et al.
Fig. 45. Control scheme of the MFGCI studied by Kuo et al. (a) Line-mode control diagram and (b) neural-mode control diagram.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 241

whose parameters are depicted in Table 33 [105]. From the control
principle diagram shown in Fig. 72(b), it can be observed that a hybrid
linear with variable structure controller (VSC) is implemented to
improve its performance. To correct the voltage unbalance at PCC
due to sudden loading or capacitor switching, the control approach in
Fig. 73 is researched.
In order to mitigate the voltage flickers of the DGSs result from
arc furnaces, arc welders, and/or starting of motors using MFGCI,
Marei et al. have researched a DGS as shown in Fig. 74 [106]. From
the control principle diagram as displayed in Fig. 75, it can be
found that Hilbert transform is employed to detect the magnitude
of the voltage at PCC.
To control the active and reactive power of a GCI flexibly, M.
Saitou and T. Shimizu have investigated a MFGCI system using
Hilbert transform, whose important parameters are listed in
Table 34 [107]. From the configuration of the MFGCI in Fig. 76,it
can be concluded that the Hilbert transform is applied to calculate
the instantaneous power and lock the phase of utility easily, which
can avoid the complex and time-consuming PLL. The reference
reactive power Q
s
can be set appropriately to satisfy the require-
ments for RPI mode or reactive injection in some special operation
circumstances.
Chandhaket et al. also give a three-phase MFGCI configuration
using soft-switching technology, as demonstrated in Fig. 77 and
Table 35 [108,109]. This topology mainly consists of PV array,
auxiliary active resonant commutated snubber link (ARCSL), and
LCL-filter. There are two operation modes for this MFGCI, as shown
in Fig. 78. When the load is small (less than 35 kW), the converter
works as a PWM rectifier, and the utility charges the battery using
the bi-directional converter. When the load is big (more than
50 kW), the converter acts as a GCI with APF functionality, which
can supply peak power and harmonic current to the load. The control
strategy of this MFGCI configuration is illustrated in Fig. 79.
From the control diagram depicted in Fig. 79, it can be found
that, due to the utility voltage is used to lock phase, the current i
q
can determine the power flow of the bi-directional converter.
When the load power P
L
is bigger than its maximum limitation
P
Lmax
, set i
n
q
>0, so the converter will supply power to load. On the
contrary, when P
L
is less than its minimum limitation P
Lmin
, set
i
n
q
<0, thus the bi-directional converter works as a PWM rectifier
to charge battery. Especially, when P
L
satisfies P
Lmin
<P
L
<P
Lmax
,an
optimal operation scheme will be carried out according to the
state-of-charge (SOC) of the battery. It should be noted that the
instantaneous power theory mentioned before is employed to
detect the harmonic current.
Prodanovic et al. have also investigated a MFGCI configuration
as presented in Fig. 80 to compensate the harmonic and reactive
current of the distributed generation system [110,111]. The basic
idea of the MFGCI is to implement the voltage, across the line
impedance, to enhance the power quality at PCC according to the
configuration in Fig. 80(a). As shown in Fig. 80(b), the Kalman
observer is employed to detect the fundamental and harmonic
components of current and voltage for compensation. Besides, a
power and voltage control sub-system is utilized to generate the
reference current for power generation tracking. As shown in
Fig. 81, three approaches are available to calculate the reactive
power control: (a) to keep the output reactive power of the MFGCI
following the desired one, (b) to track the reference voltage, and
(c) to limit the voltage in a band range.
The MFGCI configurations mentioned before mainly PV grid-
connected systems. However, wind turbine grid-connected sys-
tems can also be implanted as MFGCIs [112–114 ]. A MFGCI
configuration using DIFG is studied by Abolhassani et al. as
illustrated in Fig. 82 and Table 36. It can suppress the harmonic
current of the nonlinear load by the means of harmonic current
compensation using the stator of the DFIG, whose control strategy
Fig. 48. Block diagram of the half bridge zero-voltage switch DC/DC converter.
Fig. 49. Block diagram of the inverter stage.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270242

is demonstrated in Fig. 83. As the conventional DFIG system, the
converter in generator side controls the speed and torque of the
DFIG. However, the control of converter in utility side can attach
harmonic current compensation functionality. Besides, the refer-
ence harmonic current in d- and q-axes, i
en
qLh
and i
en
dLh
, are added
as well.
The before mentioned MFGCIs are mainly VSIs. However,
current-source-inverters (CSIs) and Z-source-inverters (ZSIs) can
also be employed as MFGCIs. Gajanayake et al. have investigated a
ZSI based MFGCI configuration as shown in Fig. 84, whose
parameters are described in Table 37. To maintain the voltage at
PCC, the utility current I
g
is controlled indirectly. Therefore, the
performance is greatly determined by the utility inductor L
g
. For a
stiff grid with a small utility inductor, perhaps the utility current
fully decouples with the voltage at PCC. In other words, the
harmonic currents across utility inductor can hardly distort the
voltage V
pcc
.
As shown in Fig. 85, its control strategy consists of a PR
controller and a time delay controller, which can satisfy the
excellent steady and dynamic performance of the MFGCI.
To maintain the power quality at PCC, the reference generator of
the controller is a very important component. An improved
Fig. 50. Three-phase H-bridge MFGCI presented by Wu et al. (a) Detailed configuration of the MFGCI and (b) equivalent two ports model of a three-phase system.
Table 22
Parameters of the three-phase full-bridge MFGCI investigated by Wu et al.
Dc-source PV array, voltage of dc-bus is 40 0 V
Capacity 1.1 kW
Voltage of utility grid 110 V/60 Hz
Switching frequency 20 kHz
Passive components C
dc
¼2200 μF, L
s
¼5mH
Control strategy PI control, SPWM modulation
Extra function APF
Fig. 51. Control block diagram of proposed inverter system (j¼uor w).
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 243

reference current generator is proposed by Gajanayake et al., as
displayed in Fig. 86 [115]. A moving window RMS calculator is
utilized to ensure the generated RMS value of the MFGCI output
current is kept free from the ripple components. A PI regulator is
implemented to generate the multiplier K
1
so as to exploit the
remaining capacity of the MFGCI as much as possible. Besides, it
can also prevent the damage to the MFGCI by the unwanted
excessive current. It should be noted that a limiter is embedded to
avoid the reference current of the MFGCI exceeding its rated one.
As previously mentioned, the MFGCIs mainly implement
H-bridge. However, multi-level topology can also be utilized.
Tsengenes and Adamidis have studied a MFGCI configuration as
shown in Fig. 87, whose parameters are given in Table 38 [116].
The dc source of the MFGCI is comprised of PV arrays in parallel.
In addition, a three-level neutral point clamped (NPC) inverter is
employed. To compensate the harmonic current of local load, the
control strategy presented in Fig. 88 is implemented. It should be
noted that the instantaneous reactive theory is utilized to generate
the compensating current components.
Note that the MFGCI topologies mentioned before can also be
suitable for other RESs application. For the energy storage devices,
for instance battery, super capacitor, and superconductivity, their
output dc terminals can substitute the dc-link of those topologies.
For the ac micro-sources, for instance direct-driven wind turbines,
gas turbines, and flywheels, there are diode rectifiers or PWM
rectifiers to connect them to the MFGCI topologies.
On the other hand, the topologies previously mentioned mainly
take three-phase H-bridge structure. Because there is no neutral
line, it can hardly compensate unbalance load current in three-
phase four-wire system. To form a neutral line, the split capacitor
in dc-bus is a solution. Therefore, the buffer capacitor is split into
two parts, and the midpoint of the capacitors is regarded as the
neural point. However, the voltages of the two capacitors are hard
to balance, which is a very important feature for this kind of
structure. To overcome this drawback, some enhanced control
strategy should be carried out. In addition, the neural line may
flow across big current, which is another disadvantage of this
structure. An advanced topology for three-phase four-wire system
application is the three-phase four-bridge topology. In this kind of
topology, the fourth bridge is employed to control neural line,
therefore it will increase the cost of system compared with three-
phase H-bridge topology. However, the fourth bridge will greatly
enhance the freedom of control strategy.
Fig. 53. Control block diagram of the MFGCI in [90].
Fig. 52. Detailed block diagram to decompose load currents into real, reactive,
and harmonic components.
Fig. 54. Configuration of the MFGCI presented by He et al. in different work mode.
(a) Current control mode and (b) voltage control mode.
Table 23
System parameters of the MFGCI presented by He et al.
Dc source DG, V
dc
¼260 V
Capacity 5 kVA
Utility voltage 104 V/60 Hz
Sampling frequency 12 kHz
Passive components DG impedance: R¼1Ω,L¼5mH
Grid impedance: R¼1Ω,L¼5mH
LC filter: L¼2.5 mH, C¼40 μF
Control strategy Spatial repetitive control, SPWM modulation
Extra functions APF
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270244

Sawant et al. investigate a MFGCI configuration to compensate
the harmonic and unbalance current of load using three-phase
four-bridge topology, as shown in Fig. 89, whose parameters are
listed in Table 39 [117]. It can be seen that a back-to-back converter
is implemented to connect a direct-driven wind turbine to utility.
And the DC/AC converter uses the topology illustrated in Fig. 89(b),
which is look forward to act as a MFGCI.
The control strategy of the MFGCI presented by Sawant et al.
can be described as shown in Fig. 90. There are two algorithms to
generate reference current in pqr frame, namely “reference power
control method (RPCM)”and “reference current control method
(RCCM)”, whose freedoms are 3 and 4, respectively. Meanwhile, a
3D-SVPWM modulation is employed to generate the trigger
pulses. In addition, Sawant et al. also gives a MFGCI configuration
using split capacitor as shown in Fig. 91 [118].
Wang et al. have presented a MFGCI configuration for compen-
sation of voltage dig and unbalance using a three-phase four-leg
topology as demonstrated in Fig. 92 and Table 40 [119,120]. From
Fig. 92, it can be seen that the sensitive loads attached at PCC may
be greatly affected by the unbalance voltage of utility. Thanks to
the inductors L
sabc
, the voltage at PCC can be can be regulated by
the MFGCI in some certain. To achieve this functionality, the
control principle of the MFGCI is presented as depicted in Fig. 93(a).
It can be found that a complex control algorithm is employed in
Fig. 56. Three-phase H-bridge MFGCI presented by Kim et al.
Fig. 55. Schematic diagram of the MFGCI system presented by Yu et al.
Table 24
Parameters of the MFGCI presented by Yu et al.
Dc source DG, v
dc
¼800 V
Utility voltage 380 V
Capacity 10 kW
Control strategy PI control, SPWM modulation
Extra functions APF and RPI
Fig. 58. Schematic diagram of the MFGCI presented by Dasgupta et al.
Fig. 57. Block diagram of photovoltaic system control.
Table 25
Parameters of the three-phase H-bridge MFGCI proposed by Kim et al.
Dc-source PV array, battery, voltage of dc-bus 260 V
Voltage of utility grid 110 V/60 Hz
Switching frequency 20 kHz
Passive components C
dc
¼3500 μF, L
f
¼7mH
Control strategy Hysteresis modulation
Extra functions PFC and UPS
Table 26
System parameters of the MFGCI presented by Dasgupta et al.
Dc source DC source, V
dc
¼200 V
Capacity 75 W
Utility voltage Approximately 40 V
Passive components L¼2.5 mH, R¼1Ω,C
dc
¼180 μF
Control strategy Lyapunov control, SPWM modulation
Extra functions APF
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 245

dual-synchronous rotating frames. Power generation tracking is
achieved in the positive-sequence frame. On the other hand, the
unbalance voltage correction is realized in the negative-sequence
frame. It should be noted that a θ-shift is utilized to form the
reference current in negative-sequence to compensate unbalance
component of the utility voltage. Besides, the active power of the
MFGCI is regulated by a PI controller. Additionally, a direction
control is embedded. Thus, if k
dir
¼−1, the MFGCI can interface the
DG to utility; whereas, the energy of utility can be fed to energy
storage DG, such as batteries, when k
dir
¼1. Furthermore, a novel
algorithm named as multi-variable filter is proposed to separate
the positive- and negative-sequence components of the unbalance
utility voltage, as proposed in Fig. 93(b).
Pinto et al. have given a MFGCI configuration as shown in
Fig. 94, whose parameters are described in Table 41 [121].
Its control strategy is given in Fig. 95, from which it can be found
that instantaneous power theory is employed to form the harmo-
nic and unbalance current for compensation.
Fig. 59. Control scheme of the MFGCI investigated by Dasgupta et al.
Fig. 60. Three-phase H-bridge topology of the MFGCI by Cheng et al.
Table 27
Parameters of the three-phase H-bridge topology
by Cheng et al.
Dc-source PV array
Capacity 5 kVA
Extra function PFC and APF
Fig. 61. Algorithm to generate reference current.
Fig. 62. Single-line diagram of the MFGCI and its control principle.
Table 28
Parameters of the MFGCI presented by Naderi et al.
Dc source DG, V
dc
¼100 0 V
Utility voltage 380 V
Passive components R
s
¼0.05 Ω,L
s
¼1mH,R
c
¼0.1 Ω
Control strategy PI control, hysteresis modulation
Extra functions APF and RPI
Fig. 63. Block diagram of the controller.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270246

Although three-phase four-bridge topology may be a good
choice for the MFGCI application in three-phase four-wire utility
system, three-phase full-bridge topology may be also suitable for
some consideration. Especially, the increasing cost due to 4 more
IGBTs may be less than the benefit due to voltage and current
reduce of IGBTs, compared with three-phase four-bridge structure.
On the other hand, the bulk transformers in three-phase full-
bridge structure can provide reliable and effective isolation, which
is greatly suitable for some situations required high reliability.
It should be note that the ac side of this topology can be decoupled,
therefore three single-phase GCI can be totally independent and
flexibly operate in parallel. Of cause, this topology can also be
decoupled in dc side, which can connect three independent micro-
sources to utility. It is particularly suitable for energy storage devices
application. Because of the dc decoupling, three energy storage
devices can be controlled independently, so the SOC of different
energy storage devices can be easy to balance. Additionally, the dc
voltage can be low, and the set-up DC/DC converter with high gain
can be cancelled.
Table 29
Parameters of the single-phase full-bridge MFGCI presented by Mohod et al.
Dc source Battery, the voltage of dc-link is V
dc
¼800 V
Capacity 150 kW
Utility voltage Three-phase 415 V/50 Hz
Passive
components
Micro-wind generator: 150 kW, 415 V, 50 Hz, 4 poles,
R
s
¼0.01 Ω,R
r
¼0.015 Ω,L
s
¼0.06 H, L
r
¼0.06 H, wind
velocity 5 m/s
Battery: dc 80 0 V, cell capacity 500 Ah, type-lead acid
Transformer: 1 kVA, Y–Y type, 415/800 V, 50 Hz
Load: three-phase nonlinear load, R¼10 Ω,C¼1μF
Power electronic
devices
IGBT: rated voltage 1200 V, forward current 50 A, gate
voltage 720 V, turn-on delay 70 ns, turn-off delay 400 ns,
power dissipation 300 W
Control strategy PI control, hysteresis modulation
Extra functions APF
Fig. 64. Configuration of the MFGCI presented by Mohod et al.
Fig. 65. Control scheme of the MFGCI studied by Mohod et al.
Fig. 66. Block diagram of the proposed MFGCI control system by Marei et al.
Table 30
Parameters of the three-phase H-bridge topology by Marei et al.
Dc-source Micro-source, voltage of dc-bus V
dc
¼500 V
Voltage of utility
grid
200 V (RMS value of line-voltage)/60 Hz
Passive
components
L
s
¼1 mH, R
s
¼0.07 Ω,L
c
¼1.6 m H, R¼1Ω,L¼3 mH,
R
b
¼0.1 Ω,L
b
¼1.68 mH, R
c
¼5Ω,L
dc
¼3.5 mH, R
dc
¼1.5~3.5 Ω
Control strategy FLC control, PI control, SPWM modulation
Extra function APF
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 247

Majumder et al. give a MFGCI configuration using three-phase
full-bridge structure, as explained in Fig. 96 [122,123]. In addition,
an ac MG application with this kind of MFGCI is studied.
For each MFGCI, to prevent the total reference current, com-
posed of power generation tracking and power quality compensa-
tion functionalities, exceeding its rated one, an algorithm to
calculate the reference current is presented. When the rated
apparent of MFGCI is larger than the demand of harmonic and
unbalance loads, it takes
P
MG
¼P
comp,rated
−P
Lav
<0
Q
MG
¼Q
comp,rated
−Q
Lav
<0
(ð11Þ
where P
MG
/Q
MG
,P
Lav
/Q
Lav
, and P
comp
/Q
comp
, are the active and
reactive power supplied by the MG, the demanded of load, and
supplied by MFGCI, respectively. Furthermore, the compensation
Fig. 67. Control diagram of the MFGCI configuration presented by Marei et al. (a) Schematic diagram of the control strategy and (b) block diagram of the detection approach
using adaptive neurons.
Fig. 68. MFGCI configuration presented by Cheng et al.
Table 31
Parameters of the three phase H-bridge topology by Cheng et al.
Dc-source Micro-source, V
dc
¼380 V
Voltage of
utility grid
220 V(RMS value of line-voltage)/60 Hz
Capacity 1 kVA
Switching
frequency
20 kHz
Passive
components
Two-phase unbalance load 46.6 Ω, three-
phase balance load 280 Ω
Control
strategy
Droop control, PI control, SPWM modulation
Extra function Unbalance compensation
Fig. 69. Droop control strategy of the MFGCI configuration proposed by Cheng et al.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270248

power supplied by the MFGCI can be expressed as
i
comp,a
i
comp,b
i
comp,c
2
6
43
7
5¼
i
NLa
i
NLb
i
NLc
2
6
43
7
5þ1
K
3P
MG
v
pa
þffiffiffi
3
pQ
MG
ðv
pb
−v
pc
Þ
3P
MG
v
pb
þffiffiffi
3
pQ
MG
ðv
pc
−v
pa
Þ
3P
MG
v
pc
þffiffiffi
3
pQ
MG
ðv
pa
−v
pb
Þ
2
6
6
43
7
7
5ð12Þ
where K¼v
2
pa
þv
2
pb
þv
2
pc
is the sum of square of each utility
phase-voltage. When the rated apparent power of the MFGCI is
less than the demand of loads, it takes
P
MG
¼P
Lav
−P
comp
¼P
Lav
−λ
P
P
Lav
¼P
Lav
ð1−λ
P
Þ>0
Q
MG
¼Q
Lav
−Q
comp
¼Q
Lav
−λ
Q
Q
Lav
¼Q
Lav
ð1−λ
Q
Þ>0
(
ð13Þ
where λ
P
∈(0, 1) and λ
Q
∈(0, 1) are compensation coefficients of the
MFGCI. It is easy to found that the MG will supply part of harmonic
and reactive current of the load, when the MFGCI cannot satisfy
the demand of loads fully. In this constitution, the reference
current of the MFGCI can be written as
i
comp,a
i
comp,b
i
comp,c
2
6
43
7
5¼
i
NLa
i
NLb
i
NLc
2
6
43
7
5−1
K
3P
Lav
ð1−λ
P
Þv
pa
þffiffiffi
3
pQ
MG
ð1−λ
Q
Þðv
pb
−v
pc
Þ
3P
Lav
ð1−λ
P
Þv
pb
þffiffiffi
3
pQ
MG
ð1−λ
Q
Þðv
pc
−v
pa
Þ
3P
Lav
ð1−λ
P
Þv
pc
þffiffiffi
3
pQ
MG
ð1−λ
Q
Þðv
pa
−v
pb
Þ
2
6
6
43
7
7
5
ð14Þ
Fig. 70. The MFGCI configuration presented by Lv et al.
Table 32
Parameters of the three phase full bridge topology by Lv et al.
Dc-source Micro-source, V
dc
¼100 0 V
Voltage of utility
grid
230 V(RMS value of phase-voltage)/50 Hz
Capacity 400 kVA
Switching frequency 12.8 kHz
Passive components Buffer capacitor C
dc
¼10 mF; LC filter L
f
¼0.2 mH,
C
f
¼30 μF
Control strategy PI control, SPWM modulation
Extra function APF
Fig. 72. Configuration of the MFGCI studied by Mohamed and El Saadany. (a) Overview diagram and (b) detailed block diagram of the MFGCI.
Fig. 71. Control strategy of the MFGCI configuration proposed by Lv et al.
Table 33
Parameters of the single-phase full-bridge MFGCI presented by Mohamed and
E.F. El Saadany.
Dc source DG, V
dc
¼600 V
Utility voltage 110 V/60 Hz
Passive
components
L
s
¼1mH,R
s
¼0.08 Ω,L¼2.5 mH, R¼1Ω; load L
1
:20kWata
lagging power factor 0.9; load L
2
: 30 kW at a lagging power
factor 0.85; switching capacitor: 20 kVAR
Switching
frequency
6.7 kHz
Control
strategy
Hybrid linear with variable-structure control, SPWM
Extra functions DVR
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 249

The control strategy of this configuration is depicted in Fig. 97.
According to the reference current computation as mentioned above,
a linear quadratic regulator (LQR) and the hysteresis modulation are
employed to generate of the trigger pulses of the IGBTs.
In [112], Majumder et al. have noticed that there may be several
MFGCIs in a MG, it is very important to exploit the coordination
control of multiple MFGCIs. As a consequence, a coordination
control strategy using communication lines is discussed in [112] as
demonstrated in Fig. 98. However, this coordination control
approach may be hardly suitable for some complex MGs, because
the communication lines may be very long and hard to expand, the
reliability and flexibility of the MGs are lowered, as well as some
excellent features, for instance plug-and-play and hot-swap,
cannot be achieved. Therefore, it is very necessary to study the
coordination control strategy without communication lines.
The MFGCI configurations mentioned before can act as APF and
compensate the harmonic and reactive current of the local load.
However, the harmonic and reactive current can also be absorbed
by passive filters. Although, the fast dynamic response and good
robustness are the advantages of APF, it cannot avoid that the APF
filters are expensive compared with passive filters. On the con-
trary, the system parameters of passive filters may be varied with
the operation condition, but they are cheap and reliable. To fully
utilize the advantages of active and passive filters, Chen et al. have
given a DGS configuration including a MFGCI and passive filters, as
exhibited in Fig. 99, whose parameters are given in Table 42 [17].
In this system, the MFGCI and the utility grid are the sources, for
the power quality conditioners, the passive filters are installed at
the terminals of loads; while the MFGCI is equipped at the PCC.
The inductors of utility are split as L
s
and L
t
, while the impedances
of loads are Z
Li
(i¼1, 2, …, 5). Note that the load 1, load 2, and load
3 are fed by passive filters in parallel, while the other loads are fed
by passive filters in series.
The control strategy of the MFGCI is introduced in Fig. 100.
The Fast Fourier Transform (FFT) is employed to detect the
fundamental component of the voltage. Simultaneously, the com-
pensation current of the equivalent load current i
abc
is identified
using instantaneous reactive power theory. It is worth nothing to
note that a SPWM modulation is also used to generate the trigger
pulses of voltage source converter (VSC). According to this control
strategy, the PMSG can inject active, reactive, and harmonic
current to the DGS. As a consequence, the utility just need absorb
or supply fundamental active current. The power flow of this
configuration is illustrated in Fig. 101.
So far, the previously mentioned MFGCIs mainly catch the extra
compensation functionalities in parallel such as APF, PFC, and UPS
and so on. To expand the extra functionalities of the MFGCIs, Han
et al. have investigated a MFGCI configuration, as depicted in
Fig. 102, whose parameters are listed in Table 43 [124]. This MFGCI
can perform the power quality condition functionalities like
an UPQC.
From the configuration in Fig. 102, it can be seen that the dc-
source of this MFGCI is connected to a distributed generator (DG),
which is different from an UPQC has the same topology.
A schematic diagram of its control principle is depicted in Fig. 103.
The conventional UPQC directly absorbs power to maintain the
voltage of the dc-link, while the dc-link of the MFGCI is fed by RESs
and/or energy storage devices. It should be noted that the series
converter works just when the utility voltage is sag, swell, and/or
unbalance; on the contrary, the parallel converter works all the time
Fig. 73. Control strategy for unbalance voltage correction.
Fig. 74. Configuration of the MFGCI studied by Marei et al.
Fig. 75. Control strategy of the MFGCI presented by Marei et al.
Table 34
Parameters of the single-phase full-bridge MFGCI
presented by Saitou and Shimizu.
Dc source Battery, V
dc
¼600 V
Utility voltage 110 V/50 Hz
Passive components L
s
¼1.2 m H
Switching frequency 15 kHz
Control strategy PI, SPWM
Extra functions RPI
Fig. 76. Configuration of the MFGCI investigated by Saitou and Shimizu.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270250

as the interface to generate active power and compensate power
quality issues.
Fig. 104 shows the configuration of the parallel converter,
which is the same as conventional three-phase GCI. For the DG
subsystem, the voltage of the DG is detected to control the
terminal voltage of DG is stable, which is fed by a transformer
and a diode rectifier. The PLL is employed to catch the phase of the
positive-sequence utility voltage, which can avoid the disturbance
from unbalance and distorted components of utility voltage. For
the series converter, its output voltage and current are sampled to
control the compensation voltage in series. The shunt converter
samples its output current, the voltage and current of loads to
supply the harmonic current to nonlinear load.
In detail, the control strategy of series converter is demon-
strated in Fig. 105. The deviation between the standard and
sampled utility voltage, namely V
ref
and V
s
, derive the reference
voltage of converter V
n
F
. Then a PI regulator is utilized to form the
reference current I
n
SF
. Due to the outer current loop, the modula-
tion signal for PWM logic V
n
C
is achieved to generate the trigger
pulses of IGBTs.
The shunt converter has two operation modes, namely APF
mode and UPS mode. In APF mode, the shunt converter acts as a
current-controlled source, whose control diagram is introduced in
Fig. 106. The detected positive-sequence voltage v
′
sαβ
and load
current i
Lαβ
in αβ frame can be utilized to calculate the instanta-
neous power pand q. It should be noted that the regulator of dc-
Fig. 77. Utility connected bi-directional soft-switching MFGCI presented by Chandhaket et al.
Table 35
Parameters of the MFGCI proposed by Chandhaket et al.
Dc-source Micro-source, V
dc
¼440 V
Capacity 20 kVA
Voltage of utility grid 200 V (RMS value of line-voltage)/60 Hz
Sampling frequency 12 kHz
parameters L
b
¼0.1 H, R
b1
¼0.05 Ω,R
b2
¼0.05 Ω,C
b
¼1F, L
f
¼3mH,R
Lf
¼0.05 Ω,C
r
¼55 μF, C
n
¼9.7 mF
Control strategy PI, space vector pulse width modulation (SVPWM)
Extra function PWM rectifier, APF
Fig. 78. Two operation modes of the MFGCI configuration presented by Chandhaket et al. (a) Energy supply (inverter) mode and (b) energy storage (rectifier) mode.
Fig. 79. Control strategy of the MFGCI configuration presented by Chandhaket et al.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 251

bus can be added to active power control loop, and instantaneous
power theory is employed to form the reference current. The detailed
block diagram of controller is explained in Fig. 107.Furthermore,a
current regulator and a forward controller are employed to form the
modulation signal. However, when the shunt converter works under
UPS mode, it acts as a voltage-controlled source, as depicted in
Fig. 107. Under such mode, an outer voltage loop and an inner current
loop are utilized, which is simpler compared to APF mode. Associated
with protection circuit, the modulation signal, and PWM logic, the
trigger pulses of IGBTs are achieved.
Similarly, Li et al. also have researched a MFGCI configuration
including three-phase four-bridge shunt and series converters, as
shown in Fig. 108, whose parameters are described in Table 44
[125,126]. When the utility grid is interrupted, the shunt converter
A acts as an UPS to supply the sensitive load. When the utility
voltage is distorted, the series converter B can compensation the
harmonic voltage components to satisfy sensitive load.
When the utility voltage is interrupted, the MFGCI acts as an
UPS and generate a stable voltage. The brief and detailed control
diagrams are exhibited in Fig. 109. It can be found that, the
deviations of power angle and voltage amplitude are used to
regulate the output active and reactive power of converter A,
which yields its reference voltage in αβ0 frame. Then complex
controllers in positive-, negative-, and zero-sequence are
employed to accurately track the reference voltage.
The control strategy of series converter is given in Fig. 110. From
the overall block diagram in Fig. 110(a), it can be seen that the
control strategy is comprised of an outer current loop and an inner
voltage loop. A detailed description is available in Fig. 110(b). PI
regulators in dq-axis and cross-decoupling terms are utilized in
the outer current loop, while a PR controller is employed in 0-axis.
The inner voltage loop is shown in Fig. 110(c). Similarly, PR
controllers are implemented, where 1/K
in
¼V
dc
/2 is the amplifica-
tion coefficient of the inverter.
It should be noted that the control strategy of shunt converter is
given in Fig . 111 to keep the terminal voltage of load, when the utility
voltagesags.Becausetheremaybelargecurrentacrossshunt
converter to damage the power electronic devices, a control approach
named as flux-charge scheme is proposed for series converter. Due to
the approach in Fig . 111 , the series converter can be viewed as a
virtual inductor to limit the possible over-current of shunt converter,
which can ride-through the utility voltage sag effectively.
Wang et al. have described a UPQC-based MFGCI configuration
as shown in Fig. 112, whose parameters are listed in Table 45 [127].
The MFGCI in Fig. 112(a) can be implemented to a micro-grid as
shown in Fig. 112(b) to enhance its power quality at PCC. A detailed
diagram of the MFGCI is shown in Fig. 112(c), as well as its control
scheme is described in Fig. 113.
According to the control principle of the parallel converter as
depicted in Fig. 113(a) and (b), PR controllers G
cαβγ
(s) are utilized, in
such a way that an excellent steady state feature can be obtained.
Besides, the transfers F
iαβγ
(s)¼K
fI
s/(s+2πf
hp
), are employed to
enhance the disturbance sensitivity of the MFGCI. Furthermore,
F
iαβγ
(s)¼K
ff
e
−Td
is transfer delay function, where K
ff
denotes the
forward gain. Similarly, the control scheme of the series converter
can be illustrated as shown in Fig. 113(c) and (d). It should be
noted that a weighted currents feedback control approach is
utilized to reduce the third-order LCL-filter model as a first-
order one [128].
Fig. 80. Configuration of the MFGCI presented by Prodanovic et al. (a) Overall system and (b) system architecture in detail.
Fig. 81. Detailed block diagram for power and voltage control.
Fig. 82. Block diagram of the proposed method.
Table 36
Parameters of the three phase full bridge topology
by Abolhassani et al.
Dc-source DFIG
Capacity 7.5 kW
Voltage of utility grid 230 V/60 Hz
Sampling frequency 20 kHz
Extra function APF
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270252

Yu et al. have given a similar MFGCI configuration, as demon-
strated in Fig. 114 [129]. It can be found that, the MFGCI and MG
share the dc-bus, while the parallel converter connects to ac-bus.
It should be noted that the series converter embeds ac lines, which
can compensate utility interruption.
The parallel converter is used to compensate the harmonic
current of nonlinear load, as well as, the simplified equivalent
circuit is exhibited in Fig. 115, while the control of parallel
converter is shown in Fig. 116.Acurrentloopisemployedto
track the fundamental and harmonic current components
injected into utility. Because the amplitude of harmonic current
and its order satisfy a hyperbolic function for diode rectifier load
(and fifth and seventh components are major part), for con-
venient, a PR controller for the sixth order harmonic is
employed to accurately compensate the fifth and seventh
harmonic components.
Fig. 117 illustrates the equivalent circuit and the control
strategy of the series converter. The series converter can compen-
sate the voltage sag or swell, by the means of the outer voltage
loop and the inner current loop. As a result, the voltage of the MG
at PCC can satisfy standards.
Fig. 83. Block diagram of the rotor side controller of the alternator/active filter for adjustable wind turbine.
Fig. 84. Schematic and detailed diagram of the MFGCI presented by Gajanayake et al. (a) Overview diagram and (b) configuration of the MFGCI using ZSI in detail.
Table 37
Parameters of the single-phase full-bridge MFGCI presented by Gajanayake et al.
Dc source The voltage of dc-link is V
dc
¼60–100 V
Capacity 1 kVA
Utility voltage RMS voltage of each phase 35 V
Passive components L
f
¼10 mH, L¼3.5 mH, C¼15 00 μF, L
g
¼1.6 m H
Control strategy PR and PI control, SPWM modulation
Extra functions APF
Fig. 85. Controller for ac side of the MFGCI.
Fig. 86. Reference current generator for power quality improvement.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 253

Fig. 87. Configuration of the MFGCI presented by Tsengenes and Adamidis.
Table 38
Parameters of the MFGCI presented by Tsengenes and Adamidis.
Dc source PV array, v
dcref
¼110 0 V
Utility voltage 230 V
Passive components R
c
¼0.1 Ω,L
c
¼0.81 mH, C
1
¼C
2
¼4mF
Control strategy PI control, SVPWM modulation
Extra functions APF
Fig. 88. Control strategy of the MFGCI proposed by Tsengenes and Adamidis.
Fig. 89. Three-phase four-leg MFGCI presented by Sawant et al. (a) System configuration and (b) three-phase four-bridge converter.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270254

Table 39
Parameters of the three-phase four-leg MFGCI proposed by Sawant et al.
Dc-source Wind turbine, voltage of dc-bus V
dc
¼800 V
Voltage of utility grid 220 V/50 Hz
Switching frequency 10 kHz
Passive components L
s
¼1.2 mH, R
s
¼10 mΩ,R
i
¼20 mΩ,L
i
¼2.4 mH (i¼a,b,c), R
n
¼20 mΩ,L
n
¼30 mH
Control strategy 3D-SVPWM modulation
Extra function APF
Fig. 90. Schematic block diagram of the MFGCI with four-leg voltage-source converter.
Fig. 91. The MFGCI configuration using split capacitor investigated by Sawant et al.
Fig. 92. Schematic diagram of the MFGCI.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 255

Table 40
Parameters of the single-phase full-bridge MFGCI presented by Wang et al.
Dc source PV array
Capacity 1 kW
Utility voltage 110 V/50 Hz
Passive
components
L
1
¼5 mH, C
f
¼4μF, L
f
¼2.5 mH; diode rectifier
L¼150 mH, R¼35 Ω
Switching
frequency
10 kHz
Control strategy PI control, SPWM
Extra functions APF
Fig. 93. Reference current generation for the MFGCI presented by Wang et al. (a) Block diagram of the control principle and (b) the novel algorithm for positive- and
negative-sequence detection.
Fig. 94. Three-phase four-leg MFGCI configuration researched by Pinto et al. (a) System configuration and (b) interface system and p–qtheory power components.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270256

When the utility grid is interrupted, the MG transfers to
islanded mode. Meanwhile the parallel converter acts as an UPS
and supply reference voltage for other GCIs and loads in the MG to
keep the MG stable.
Table 41
Parameters of the three-phase four-leg MFGCI presented by Pinto et al.
Voltage of utility grid 75 V/50 Hz
Switching frequency 10 kHz
Passive components Unbalance loads: 20 0 mH, 200 mH, 0 mH Nonlinear loads: 60 Ω,68mH
Power electronic devices IGBT
Extra function APF and unbalance compensation
Fig. 95. Control strategy of the MFGCI configuration proposed by Pinto et al.
Fig. 96. MG and utility system under consideration by Majumder et al. (a) Micro-grid with two MFGCIs, (b) configuration of the MFGCI and (c) the power flow of the system.
Fig. 97. Control strategy of the MFGCI topology presented by Majumder et al.
Fig. 98. A MG structure with several MFGCIs and loads.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 257

6. Analysis and discussion
MFGCIs are special GCIs, whose control strategies can partly
inherit the ones of conventional GCIs. MFGCIS can be classified as
many categories according to different considerations, as exhibited
in Fig. 118.
According to the utility, MFGCIs can be classified as depicted
in Fig. 118(a). It can be seen that single-phase and three-
phase utility application drawn more attention; whereas,
there has no MFGCI for two-phase system application. However,
two-phase utility grids are widely implemented in railway sys-
tems. Thus, this research point maybe encouraging in the
following years.
From the view of modulation in Fig. 118(b), the control
strategies of MFGCIs can be classified as three categories, namely
hysteresis, SPWM, and SVPWM [130,131]. Among them, the
hysteresis modulation has the advantage of fast dynamic response.
However, the switching frequency is not constant in hysteresis
modulation, and some improved hysteresis modulation approaches
with constant switching frequency are very complex [132]. Therefore
it may burden the filter and controller design. On the contrary, the
SPWM modulation has gained common attention due to its constant
switching frequency and flexible control approaches. For instance,
PI control [133], PR control [134], weighted currents feedback
control [128], deadbeat control [135], repetitive control [136,137],
iterative learning control [138], and robust control [139] all can be
implemented on SPWM modulation. However, the obvious drawback
of SPWM modulation is the low efficiency of dc voltage. Besides,
SVPWM is good modulation due to its high efficiency of dc voltage,
which utilizes eight space voltage vectors to approximately emulate
the rotating voltage vector. It also has the advantages of constant
switching frequency and flexible control approaches. In general, the
control approaches suitable for the SPWM can be implemented on
Fig. 99. Configuration of a DGS with MFGCI presented by Chen et al.
Table 42
Parameters of the system studied by Chen et al.
Dc-source PMG (permanent magnet generator), voltage of dc-bus V
dc
¼960 V
Switching
frequency
3.15 kHz
Passive
components
Load 1 is a thyristor rectifier in three-phase, whose trigger angle is 581. The active and reactive power of passive filter PF 1 are P¼290 kW
and Q¼−457 kW
Load 2 is a diode rectifier in single-phase. The active and reactive power of passive filter PF 2 are P¼−134 kW and Q¼−167 kW, whose unbalance
coefficient is 8.2%
Load 3 is a thyristor rectifier in three-phase, whose trigger angle is 101. The active and reactive power of passive filter PF 3 are P¼−110 kW and
Q¼−8kW
Load 4 is a diode rectifier in three-phase. The active and reactive power of passive filter LCL 4 are P¼−60 kW and Q¼−0.4 kW
Load 5 is a diode rectifier in three-phase. The active and reactive power of passive filter LCL 5 are P¼−30 kW and Q¼75 kW
The system inductors are L
s
¼0.01 mH and L
t
¼0.16 mH
Control strategy SPWM modulation
Extra functions APF, unbalance and harmonic current compensation using passive filters
Fig. 100. Block diagram of the control system for the VSC with FFT for functional voltage extraction.
Fig. 101. Possible instantaneous power relation in the studied system.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270258

SVPWM modulation. However, SVPWM modulation is more compli-
cated to achieve on DSP than SPWM.
From the view of auxiliary functionalities of MFGCIs, they
can be classified as illustrated in Fig. 118(c). Because the GCIs are
usual CC-VSIs, the ancillary services of MFGCIs to enhance power
quality on current issues can easily be embedded in. However, the
functionalities for voltage issues are hardly achieved using the CC-
VSIs GCIs. Thus, some special topologies are expected. As men-
tioned before, the MFGCIs can act as APF, RPI, PFC, unbalance
compensation (UC), DVR, harmonic voltage compensation
Fig. 102. Three-phase H-bridge MFGCI used as UPQC presented by Han et al. (a) Overview block diagram and (b) inverter stage of the MFGCI.
Table 43
Parameters of the MFGCI presented by Han et al.
Dc-source DG, voltage of dc-bus V
dc
¼700 V
Voltage of utility grid 220 V/60 Hz
Switching frequency 10 kHz
Passive components Buffer capacitor C
1
¼C
2
¼6600 μF
Parallel converter, filter L
f
¼600 μH, C
f
¼40 μF, switching frequency 10 kHz
Series converter, filter L
f
¼600 μH, C
f
¼40 μF, switching frequency 10 kHz, the rounding ratio of transformer 500:100, capacity 6 kVA
Capacity of nonlinear load and linear load are 17.54 kVA and 3.27 kVA, respectively
DG, capacity 30 kW, rounding ratio of transformer 380:50 0 V, voltage of diode rectifier 700 V
Inductor of system R¼1mΩ,L¼0.01 mH
Control strategy PI control
Extra function UPQC (APF, voltage sag/swell/interruption compensation)
Fig. 103. Brief block diagram of the control strategy for the MFGCI configuration be Han et al.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 259

(HVC), UPS, Voltage unbalance/interruption/sag/swell compensa-
tion (UISWC), UPQC and so on.
From the view of detection approaches of compensation
components, MFGCIs can be classified as shown in Fig. 118(d).
The auxiliary functionalities of MFGCIs are the essentially different
from conventional GCIs, where the detection approaches of
compensation components are a very important part of the control
strategies of MFGCIs. A detail comparison of different approaches
for compensation components detection is available in Table 46.
These approaches are suitable for MFGCIs application in different
conditions. The available approaches can be classified as two
categories, namely frequency domain approaches and time
domain approaches [141–143]. DFT-based algorithms are a typical
frequency domain approaches, but these kinds of methods are
complex and poor in dynamic response [144]. Therefore, the
detection approaches in time domain gain more and more attention.
These kinds of methods include FBD power theory method
[145,146], instantaneous power theory (IP) method [147,148], i
d
–i
q
method [149],i
p
–i
q
method [147], projection method [147], adaptive
filter (AF) method [150], Kalman filter method [151],neuralnetwork
(NN) method [152] and so on. It should be noted that the previously
mentioned MFGCIs mainly utilized instantaneous power theory to
detect the harmonic and reactive current of load for compensating,
because it has clear physical meaning and is easy implementation
on DSP.
Additionally, from the view of the objective for control, the
control strategies of MFGCIs can also be divided into two
categories, namely direct current control and indirect current
control [140]. The indirect current control can control the grid-
connected current by the means of voltage control. Its dynamic
response is fast, but it is sensitive to system parameters, and
the control approaches are inflexible. Therefore, the direct
current control is paid more expectation.
The MFGCI cannot only achieve power generation tracking,
but also can complete the reactive, unbalance, and harmonic
current compensation. To facilitate the algorithm of reference
current generation, MFGCIs mainly take direct current control.
However, from the view of modulation, all three kinds of
modulation methods have been used, according to the history
of modulation technology. Tabl e 47 exhibits a detailed compar-
ison of different MFGCIs topologies.
From Table 47, it can be found that, the control methods of
available MFGCIs mainly utilize direct current PI control associated
with SPWM modulation. As mentioned before, the hysteresis
modulation has the drawbacks of varied switching frequency,
which is not easy for filter design, and may lead to large current
THD. In addition, the SVPWM modulation can enhance the
efficiency of dc voltage, but it may burden the controller.
As previously mentioned, this paper has investigated the available
MFGCI topologies in capacity, switching frequency, auxiliary func-
tionalities, and application aspects in detail. There are interesting
conclusion can be drawn:

Firstly, the available MFGCIs are mainly experimental proto-
types, whose capacities are low, in general.
Fig. 104. Shunt inverter control block diagram.
Fig. 105. Series inverter control block diagram.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270260


Secondly, the auxiliary functionalities of MFGCIs still need
exploit. The current compensation mainly focuses on harmonic,
reactive, and unbalance components. Besides, the voltage
compensation mainly focuses on voltage sag/swell/
interruption. Harmonic and unbalance compensation of utility
voltage need further investigation.

Thirdly, the capacities of MFGCIs in single-phase are small,
which are mainly implemented in PV grid-connected systems.
Fig. 106. Positive-sequence detector and voltage reference generator.
Fig. 107. Control strategy of the shunt inverter. (a) Overview of the control principle and (b) the voltage control of the shunt inverter.
Fig. 108. Configuration of the MFGCI presented by Li et al. (a) Structure of the UPQC-based MFGCI and (b) power electronic topology of the shunt and series inverter.
Table 44
Parameters of the MFGCI topology by Li et al.
Dc-source Micro-sources, voltage of dc-bus 700 V
Voltage of utility grid 100 V (amplitude of phase-voltage)/50 Hz
Switching frequency 10 kHz (sampling frequency, 5 kHz, controller dSPACE DS1103-TMS320F240)
Control strategy PI control, proportional resonant control, SPWM modulation
Extra functions Voltage interruption/sag/harmonic compensation
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 261

Fig. 109. Control scheme for shunt inverter A. (a) Overall control structure and (b) voltage control algorithm.
Fig. 110. Control scheme for series inverter B. (a) General representation, (b) outer current loop in the negative synchronous frame and (c) inner voltage loop.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270262

However, the capacities of MFGCIs in three-phase are much
larger usually, which are utilized in middle- and large-scale
wind and solar plants.

At last, the switching frequency of small capacity MFGCIs are
much higher than the ones of large-capacity MFGCIs. Mean-
while, soft-switching approaches are important means to
enhance the efficiency of MFGCIs.
As mentioned before, there have many topologies and control
strategies of MFGCIs been well documented for different
capacities, application fields, and auxiliary functionalities, as well
as a new research field is exploited. However, the capacities of
existing MFGCIs are generally small and the auxiliary functional-
ities are still not perfect. Besides, it is hard to say which topology is
better than the others, and a further work on the topology theory
of MFGCIs is essential necessary. There may be some work frames
on MFGCIs researches as follows:
1. New power electronic topologies of MFGCIs. It is important to
build a uniform MFGCI configuration, which can compensate
harmonic, reactive, and unbalance current in parallel, as well can
compensate harmonic, unbalance, and sag/swell/interruption vol-
tageatthesametime.Besides,thedcvoltageofmicro-sources
should be high enough to connect to DC/AC stages of existing
MFGCIs. Therefore, a high set-up DC/DC stage may be needed,
which will increase the cost and reduce the efficiency of the
system.Insummary,newpowerelectronictopologiesofMFGCIs
should be an encouraging research field.
2. The application of MFGCIs for industrial power electric system.
The capacity of existing MFGCIs is small, and it should promote
the experimental prototype for industrial application. Simulta-
neously, some multi-level topologies, structures in parallel and/
or series should be employed to enhance the current and
voltage capacity of MFGCIs.
3. Soft-switching technology and efficiency enhancement.
The power loss and heat are very important issues for the
Fig. 111. Flux-charge control scheme for series inverter B.
Fig. 112. Configuration of the MFGCI presented by Wang et al. (a) Basic schematic diagram, (b) a micro-grid consists of the MFGCI and (c) a detailed diagram of the MFGCI.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 263

Table 45
Parameters of the single-phase full-bridge MFGCI presented by Wang et al.
Dc source The voltage of dc-link is V
dc
¼750 V
Utility voltage 230 V/50 Hz
Passive
components
Z
g
¼2 mH, L¼1.8 mH, L
n
¼0.67 mH, C¼4400 μF,
L
g
¼1.6 mH, T
n
¼1:1
Switching
frequency
16 kHz
Control strategy PR and PI control, SPWM modulation
Extra functions UPQC
Fig. 113. Scheme diagram of the MFGCI presented by Wang et al. (a) Overview of the controller of parallel converter, (b) diagram of the controller of parallel converter
in detail, (c) schematic diagram of the controller of series converter and (d) detailed block diagram of the controller of series converter.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270264

reliability and efficiency of a MFGCI. The soft-switching technol-
ogy can greatly improve these features. However, the existing
MFGCIs seldom consider such issues.
4. Novel control strategies should be exploited. As the develop-
ment of control theory and technology, the control strategies of
MFGCIs have covered hysteresis, SPWM, and SVPWM modula-
tion; however, the commonly used approach is SPWM mod-
ulation with PI controller. To obtain better performance on
steady and dynamic operation of MFGCIs, some advanced
control strategies such as LQR, robust control, and feedback
linearization control should be discussed. Besides, a DGS and
MG may contain a lot of MFGCIs, so the coordination control of
MFGCIs is a very significant scenario.
5. The stability of MFGCIs in a DGS or MG. There may be many
MFGCIs and conventional GCIs in a DGS or MG, which might
weaken the stability performance of DGS or MG to immunize
Fig. 115. Circuit model and control strategy of the MFGCI studied by Yu et al. (a) The equivalent circuit of the whole system under normal operation condition and (b) control
scheme of MFGCI.
Fig. 116. Equivalent circuit of the whole system under EPS sag/swell condition and corresponding cascaded voltage. (a) Circuit model and (b) control principle.
Fig. 114. MFGCI configuration presented by Yu et al.
Fig. 117. Control scheme for the shunt connected converter. (a) The equivalent circuit and (b) the control scheme under islanded mode.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 265

Fig. 118. Several categories of MFGCIs in different considerations. (a) Utility-based classification of MFGCIs, (b) classification of MFGCIs based on modulation and control
approaches, (c) classification of MFGCIs based on auxiliary services and (d) classification of MFGCIs based on the methods of compensation components detection.
Table 46
Detailed comparisons of different approaches to detect compensation current.
Algorithm Frequency domain Time domain
DFT IP i
d
–i
q
i
p
–i
q
Projection FBD AF Kalman NN
Easy or not √√√√ √ 
Single-phase application or not √    √√ √
Need PLL or not √  
Table 47
Comparisons of multi-functional grid-inverter topologies.
Utility Author Topology Current mode Modulation/
control
Capacity Switching
frequency (kHz)
Extra functions Application
Single-phase Kuo et al. [56] Full-bridge Direct SPWM/PI ≤1.5 kVA 20 APF PV
Wu et al. [57] Full-bridge Direct SPWM/PI ≤1.5 kVA 20 APF PV
Wu et al. [62] Full-bridge Direct SPWM/PI 1 kVA 19.45 APF, PFC PV
Sladic et al. [63] Full-bridge Direct Hysteresis –15 APF PV
Calleja and Jimenez
[65]
Full-bridge Direct Hysteresis 1 kVA 14.2 APF, RPI PV
Seo et al. [66] Full-bridge Direct SPWM/PI 3 kVA 20 APF PV
Wu and Shen [67] Full-bridge Direct SPWM/PI 1 kVA 25 APF PV
Wu et al. [68] Half-bridge Direct SPWM/PI ≤1.5 kVA 20 APF PV
Patidar et al. [70] Full-bridge Direct Hysteresis/PI 1.2 kVA 25 APF PV
Hirachi et al. [71] Full-bridge Direct SPWM/PI 3 kVA –APF PV
Dasgupta et al. [72] Full-bridge Direct SPWM/Lyapunov –10 APF Micro-source
Chiang et al. [73] Full-bridge Direct SPWM/PI ≤1 kVA –APF, UPS PV
Bojoi et al. [74] Full-bridge Direct SPWM/repetitive 4 kVA 10 APF, PFC Micro-source
Cirrincione et al. [76] Full-bridge Direct SPWM/PR –15 APF PV
Macken et al. [77] Full-bridge Direct SPWM/PI 1 kVA –APF PV
Hosseini et al. [78] Two-boost Indirect SPWM/PI ≤3 kVA 20 DVR, PFC PV
Mastromauro et al.
[79]
Full-bridge Direct SPWM/repetitive 1.2 kVA 20 DVR, HVC PV
Dasgupta et al. [82] Full-bridge Indirect SPWM/repetitive –10 DVR, HVC PV
Lin and Yang [83] Three-leg Direct SPWM/PI 1.5 kVA 20 UPQC PV
Kuo. [85] Three-leg Direct SPWM/PI 1 kVA 18 APF PV
Souza et al. [86] HB ZVS Direct SPWM/PI 1 kVA 100/10 APF PV
Three-phase Wu et al. [89] H-bridge Direct SPWM/PI 1.1 kVA 20 APF PV
He et al. [91] Full-bridge Direct SPWM/repetitive 5 kVA –APF Micro-source
Yu et al. [93] H-bridge Direct SPWM/PI 10 kVA –APF, RPI Micro-source
Kim et al. [94] H-bridge Direct Hysteresis –20 APF PV
Dasgupta et al. [95] H-bridge Direct SPWM/Lyapunov 75 VA –APF Micro-source
Cheng et al. [97] H-bridge ––5 kVA –PFC, UPS PV
Naderi et al. [98] H-bridge Direct Hysteresis ––APF, RPI Micro-source
H-bridge Direct Hysteresis 150 kVA –APF Battery
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270266

disturbance. How to analyze, judge, and control the stability of DGS
or MG need further work.
7. Conclusion
Recently, GCIs have caught common attention as important
components of DGSs and MGs, due to the deeply research on
DGS and MG to make better use of RESs. To enhance the cost-
effectiveofGCIsandthepowerqualityofDGSsandMGs,anew
and encouraging fieldonMFGCIsisexploited.Inthispaper,a
comprehensive review on the topologies and control strategies
of MFGCIs are achieved. Additionally, detailed analysis, compar-
ison, and discussion on the existing MFGCIs are investigated.
Besides, some interesting frames for further work are summar-
ized. It is expected that this review will be a helpful reference on
MFGCIs for the researchers, engineers, manufacturers, and users
concerning GCIs.
Acknowledgments
The financial support from National Natural Science Foundation
of China (No.50907060) and China Postdoctoral Science Founda-
tion funded project (20090451438) are gratefully acknowledged.
References
[1] Andersson G, Donalek P, Farmer R, Hatziargyriou N, Kamwa I, Kundur P,
Martins N, Paserba J, Pourbeik P, Sanchez Gasca J, Schulz R, Stankovic A,
Taylor C, Vittal V. Causes of the 2003 major grid blackouts in North America
Europe, and recommended means to improve system dynamic performance.
IEEE Transactions on Power Systems 2005;20(4):1922–8.
[2] Chen QQ, Yin XG, You DH, Hou H, Tong GY, Wang B, Liu H. Review on
blackout process in China Southern area main power grid in 2008 snow
disaster. In: Proceedings of the IEEE power and energy society general
meeting. Calgary, Alberta, Canada; 26–30 July, 2009. p. 1–8.
[3] Su S, Li YH, Duan XZ. Self-organized criticality of power system faults and its
application in adaptation to extreme climate. Chinese Science Bulletin
2009;54(7):1251–9.
[4] Rahman S. Green power: what is it and where can we find it? IEEE Power and
Energy Magazine 2003;1(1):30–7.
[5] Beér JM. Low carbon emission electricity generating technology options.
〈http://web.mit.edu/mitei/docs/seminars/iap10/low-carbon.pdf〉[26.12.2010].
[6] Dugan RC, McDermott TE. Distributed generation. IEEE Industry Applications
Magazine 2002;8(2):19–25.
[7] Carley S. Distributed generation: an empirical analysis of primary motivators.
Energy Policy 2009;37(5):1648–59.
[8] Amor MB, Lesage P, Pineau PO, Samson R. Can distributed generation offer
substantial benefits in a Northeastern American context? A case study of
small-scale renewable technologies using a life cycle methodology Renew-
able and Sustainable Energy Reviews 2010;14(9):2885–95.
[9] Lasseter RH. Microgrids and distributed generation. Journal of Energy
Engineering 2007;133(3):144–9.
[10] Kroposki B, Lasseter R, Ise T, Morozumi S, Papathanassiou S, Hatziargyriou N.
Making microgrids work. IEEE Power and Energy Magazine 2008;6(3):40–53.
[11] Huang JY, Jiang CW, Xu R. A review on distributed energy resources and
MicroGrid. Renewable and Sustainable Energy Reviews 2008;12(9):2472–83.
[12] Lidula NWA, Rajapakse AD. Microgrids research: a review of experimental
microgrids and test systems. Renewable and Sustainable Energy Reviews
2011;15(1):186–202.
[13] Ustun TS, Ozansoy C, Zayegh A. Recent developments in microgrids and
example cases around the world—a review. Renewable and Sustainable
Energy Reviews 2011;15(8):4030–41.
[14] Zeng Z, Yang H, Zhao R. Study on small signal stability of microgrids: a
review and a new approach. Renewable and Sustainable Energy Reviews
2011;15(9):4818–28.
[15] Blaabjerg F, Zhe C, Kjaer SB. Power electronics as efficient interface in
dispersed power generation systems. IEEE Transactions on Power Electronics
2004;19(5):1184–94.
[16] Xue YS, Chang LC, Sren BK, Bordonau J, Shimizu T. Topologies of single-phase
inverters for small distributed power generators: an overview. IEEE Transac-
tions on Power Electronics 2004;19(5):1305–14.
[17] Chen Z, Blaabjerg F, Pedersen JK. A multi-functional power electronic
converter in distributed generation power systems. In: Proceedings of the
IEEE annual power electronics specialists conference. Recife, Brazil; 12–18
June, 2005. p. 1738–44.
[18] Jardan RK, Nagy I. Power quality conditioning in distributed generation
systems. In: Proceedings of the IEEE power electronics and motion control
conference. Shanghai, China; 14–16 August 20 06. p. 1–5.
Table 47 (continued)
Utility Author Topology Current mode Modulation/
control
Capacity Switching
frequency (kHz)
Extra functions Application
Mohod and Aware
[99]
Marei et al. [100] H-bridge Direct SPWM/FLC, PI ––APF Micro-source
Cheng et al. [102] H-bridge Direct SPWM/droop
control
1 kVA 20 UC Micro-source
Lv et al. [103] H-bridge Direct SPWM/PI 40 0 kVA 12.8 APF Micro-source
Mohamed and
Saadany [105]
H-bridge Indirect SPWM/VSC –6.7 DVR Micro-source
Saitou and Shimizu
[107]
H-bridge Direct SPWM/PI –15 RPI Battery
Chandhaket et al.
[108]
H-bridge Direct SPWM/PI 20 kVA –PWM rectifier,
APF
Battery
Abolhassani et al.
[112]
H-bridge Direct SPWM/PI 7.5 kVA –APF DFIG
Gajanayake et al.
[115]
ZVI Direct SVPWM/PI 1 kVA –APF Micro-source
Tsengenes and
Adamidis [116]
Three-level NPC Direct SVPWM/PI ––APF PV
Sawant and
Chandorkar [117]
Four-bridge Direct 3D-SVPWM –10 APF, UC PMSG
Wang et al. [119] Four-bridge Direct SPWM/PI 1 kVA 10 APF PV
Pinto et al. [121] Four-bridge Direct –––APF, UC Micro-source
Majumder et al. [122] Full-bridge Direct Hysteresis/LQR ––APF, UC Micro-source
Chen et al. [17] H-bridge Direct SPWM/PI –3.15 APF, UC PMSG
Han et al. [124] H-bridge Direct SPWM/PI 30 kVA 10 APF, ISWC WT
Li et al. [125] Four-bridge Direct SPWM/PI –10 ISWC Micro-source
Wang et al. [127] Four-bridge Direct SPWM/PI, PR –16 UPQC Micro-source
Yu and
Khambadkone [129]
Four-bridge Direct SPWM/PI ––ISWC, APF Micro-source
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 267

[19] Belenguer E, Beltran H, Aparicio N. Distributed generation power inverters as
shunt active power filters for losses minimization in the distribution net-
work. In: Proceedings of the European conference on power electronics and
applications. Aalborg, Denmark; 2–5 September, 2007. p. 1–10.
[20] Macken KJP. Control of inverter-based distributed generation used to provide
premium power quality. In: Proceedings of the IEEE power electronics
specialists conference. Aachen, Germany; 20–25 June, 2004. p. 3188–94.
[21] Joos G, Ooi BT, McGillis D, Galiana FD, Marceau R. The potential of distributed
generation to provide ancillary services. In: Proceedings of the IEEE power
engineering society summer meeting. Seattle (WA), United states; 16–20
July, 2000. p. 1762–7.
[22] Triggianese M, Liccardo F, Marino P. Ancillary services performed by distributed
generation in grid integration. In: Proceedings of the international conference
on clean electrical power. Capri, Italy; 21–23 May, 2007. p. 164–70.
[23] Tsengenes G, Adamidis G. Investigation of the behavior of a three phase grid-
connected photovoltaic system to control active and reactive power. Electric
Power Systems Research 2011;81(1):177–84.
[24] Takigawa K, Okada N, Kuwabara N, Kitamura A, Yamamoto F. Development
and performance test of smart power conditioner for value-added PV
application. Solar Energy Materials and Solar Cells 2003;75(3–4):547–55.
[25] Albuquerque FL, Moraes AJ, Guimaraes GC, Sanhueza SMR,Vaz AR. Photovoltaic
solar system connected to the electric power grid operating as active power
generator and reactive power compensator. Solar Energy 2010;84(7):1310–7.
[26] Khadem SK, Basu M, Conlon MF. Power quality in grid connected renewable
energy systems role of custom power devices. In: Proceedings of the
international conference on renewable energy and power quality (ICREPQ).
Granada, Spain; 23–25 March, 2010. p. 1–6.
[27] Chakraborty S, Krame r B, Kroposki B. A review of power elec tronics interfaces
for distributed energy systems towards achieving low-cost modular design.
Renewable and Sustainable Energy Reviews 2009;13(9):2323–35.
[28] Kroposki B, Pink C, DeBlasio R, Thomas H, Simo X, Es M, Sen PK. Benefits of
power electronic interfaces for distributed energy systems. IEEE Transactions
on Energy Conversion 2010;25(3):901–8.
[29] Kirubakaran A, Jain S, Nema RK. A review on fuel cell technologies and power
electronic interface. Renewable and Sustainable Energy Reviews 2009;13
(9):2430–40.
[30] Zhao B, Yu Q, Sun W. Extended-phase-shift control of isolated bi-directional
DC–DC converter for power distribution in microgrid. IEEE Transactions on
Power Electronics; in press.
[31] Du Y, Lukic S, Jacobson B, Huang A. Review of high power isolated bi-
directional DC-DC converters for PHEV/EV DC charging infrastructure. In:
Proceedings of the IEEE energy conversion congress and exposition (ECCE).
Phoenix (AZ); 17–22 September, 2011. p. 553–60.
[32] Du Y, Zhou X, Bai S, Lukic S, Huang A. Review of non-isolated bi-directional
DC–DC converters for plug-in hybrid electric vehicle charge station applica-
tion at municipal parking decks. In: Proceedings of the IEEE applied power
electronics conference and exposition. Palm Springs (CA), United States; 21–
25 Febraury, 2010. p. 1145–51.
[33] Cobben JFG, Kling WL, Myrzik JMA. Power quality aspects of a future micro
grid. In: Proceedings of the IEEE international conference on future power
systems. Amsterdam, Netherlands; 18 November 2005. p. 1–5.
[34] Jonasson I, Soder L. Power quality on ships—a questionnaire evaluation
concerning island power system. In: Proceedings of the IEEE power engi-
neering society transmission and distribution conference. Vancouver, BC,
Canada; 15–19 July, 2001. p. 216–21.
[35] Enslin JHR, Heskes PJM. Harmonic interaction between a large number of
distributed power inverters and the distribution network. IEEE Transactions
on Power Electronics 2004;19(6):1586–93.
[36] Bollen MH, Hassan F. Integration of distributed generation in the power
system. Wiley-IEEE Press; 2011.
[37] Salim RH, Oleskovicz M, Ramos RA. Power quality of distributed generation
systems as affected by electromechanical oscillations—definitions and possible
solutions. IET Generation, Transmission and Distribution 2011;5(11):1114–23.
[38] Li G, Zhang Z, Li X, Wang S, Zhou M. A methodology for power quality
evaluation in distribution network with distributed generation. In: Proceed-
ings of the IEEE international conference on critical infrastructure. Beijing,
China; 20–22 September, 2010. p. 1–6.
[39] Salarvand A, Mirzaeian B, Moallem M. Obtaining a quantitative index for
power quality evaluation in competitive electricity market. IET Generation,
Transmission and Distribution 2010;4(7):810–23.
[40] Driesen J, Green T, Van Craenenbroeck T, Belmans R. The development of
power quality markets. In: Proceedings of the IEEE power engineering
society transmission and distribution conference. New York (NY), United
states; 27–31 January, 2002. p. 262–7.
[41] Hu JB, Zhang W, Wang HS, He YK, Xu L. Proportional integral plus multi-
frequency resonant current controller for grid-connected voltage source
converter under imbalanced and distorted supply voltage conditions. Journal
of Zhejiang University—Science A 2009;10(10):1532–40.
[42] Rodriguez P, Timbus AV, Teodorescu R, Liserre M, Blaabjerg F. Flexible active
power control of distributed power generation systems during grid faults.
IEEE Transactions on Industrial Electronics 2007;54(5):2583–92.
[43] Roscoe AJ, Finney SJ, Burt GM. Tradeoffs between AC power quality and DC
bus ripple for 3-phase 3-wire inverter-connected devices within microgrids.
IEEE Transactions on Power Electronics 2011;26(3):674–88.
[44] Hu Y, Chen Z, Excell P. Power quality improvement of unbalanced power
system with distributed generation units. In: Proceedings of the IEEE
international conference on electric utility deregulation and restructuring
and power technologies. Weihai, Shandong, China; 6–9 July, 2011. p. 417–23.
[45] Peng FZ, Akagi H, Nabae A. A new approach to harmonic compensation in
power systems—a combined system of shunt passive and series active filters.
IEEE Transactions on Industry Applications 1990;26(6):983–90.
[46] Akagi H, Fujita H, Wada K. Shunt active filter based on voltage detection for
harmonic termination of a radial power distribution line. IEEE Transactions
on Industry Applications 1999;35(3):638–45.
[47] Fujita H, Akagi H. Voltage-regulation performance of a shunt active filter
intended for installation on a power distribution system. IEEE Transactions
on Power Electronics 2007;22(3):1046–53.
[48] Lee TL, Li JC, Cheng PT. Discrete frequency tuning active filter for power
system harmonics. IEEE Transactions on Power Electronics 2009;24
(5):1209–17.
[49] Ghosh A, Ledwich G. Compensation of distribution system voltage using
DVR. IEEE Transactions on Power Delivery 2002;17(4):1030–6.
[50] Mahdianpoor FM, Hooshmand RA, Ataei M. A new approach to multi-
functional dynamic voltage restorer implementation for emergency control
in distribution systems. IEEE Transactions on Power Delivery 2011;26
(2):882–90.
[51] Arulampalam A, Barnes M, Jenkins N, Ekanayake JB. Power quality and
stability improvement of a wind farm using STATCOM supported with hybrid
battery energy storage. IEE Proceedings—Generation, Transmission and
Distribution 2006;153(6):701–10.
[52] Lowenstein MZ. Improving power factor in the presence of harmonics using
low-voltage tuned filters. IEEE Transactions on Industry Applications
1993;29(3):528–35.
[53] Bansal RC. Automatic reactive-power control of isolated wind-diesel hybrid
power systems. IEEE Transactions on Industrial Electronics 2006;53
(4):1116–26.
[54] Ke Z, Jiang W, Lv Z, Luo A, Kang Z. A micro-grid reactive voltage collaborative
control system configuring DSTATCOM. In: Proceedings of the IEEE mechanic
automation and control engineering. Inner Mongolia, China; 15–17 July,
2011. p. 1887–90.
[55] Sannino A, Svensson J, Larsson T. Power-electronic solutions to power quality
problems. Electric Power Systems Research 2003;66(1):71–82.
[56] Kuo Y, Liang T, Chen J. Novel maximum-power-point-tracking controller for
photovoltaic energy conversion system. IEEE Transactions on Industrial
Electronics 2001;48(3):594–601.
[57] Wu TF, Nei HS, Shen CL, Li GF. A single-phase two-wire grid-connection PV
inverter with active power filtering and nonlinear inductance consideration.
In: Proceedings of the IEEE applied power electronics conference and
exposition. Anaheim, California; 22–26 February, 2004. p. 1566–71.
[58] Wu TF, Nei HH, Shen CL, Chen TM. A single-phase inverter system for PV
power injection and active power filtering with nonlinear inductor con-
sideration. IEEE Transactions on Industry Applications 2005;41(4):1075–83.
[59] Esram T, Chapman PL. Comparison of photovoltaic array maximum power
point tracking techniques. IEEE Transactions on Energy Conversion 2007;22
(2):439–49.
[60] Esram T, Kimball JW, Krein PT, Chapman PL, Midya P. Dynamic maximum
power point tracking of photovoltaic arrays using ripple correlation control.
IEEE Transactions on Power Electronics 2006;21(5):1282–90.
[61] Salas V, Olias E, Barrado A, Lazaro A. Review of the maximum power point
tracking algorithms for stand-alone photovoltaic systems. Solar Energy
Materials and Solar Cells 2006;90(11):1555–78.
[62] Wu TF, Chang CH, Chen YK. A multi-function photovoltaic power supply
system with grid-connection and power factor correction features. In:
Proceedings of the IEEE power electronics specialists conference. Galway,
Ireland; 18–23 June, 200 0. p. 1185–90.
[63] Sladic S, Bariic B, Sokovic M. Cost-effective power converter for thin film
solar cell technology and improved power quality. Journal of Materials
Processing Technology 2008;201(1–3):786–90.
[64] Kirawanich P, O'Connell RM. Fuzzy logic control of an active power
line conditioner. IEEE Transactions on Power Electronics 2004;19(6):1574–85.
[65] Calleja H, Jimenez H. Performance of a grid connected PV system used as
active filter. Energy Conversion and Management 2004;45(15–16):2417–28.
[66] Seo HR, Jang SJ, Kim GH, Park M, Yu IK. Hardware based performance
analysis of a multi-function single-phase PV-AF system. In: Proceedings of
the IEEE energy conversion congress and exposition. San Jose, California,
USA; 20–24 September, 2009. p. 2213–7.
[67] Wu Y, Shen C. Implementation of a DC power system with PV grid-
connection and active power filtering. In: Proceedings of the IEEE power
electronics for distributed generation systems. Hefei, China; 16–18 June
2010 . p. 116–21.
[68] Wu TF, Nien HS, Hui HM, Shen CL. PV power injection and active power
filtering with amplitude-clamping and amplitude-scaling algorithms. IEEE
Transactions on Industry Applications 2007;43(3):731–41.
[69] Wu TF, Nien HS, Shen CL, Hsieh HM Chen YM. A half-bridge 1Φ2W PV
inverter system with active power filtering and real power injection. In:
Proceedings of the IEEE applied power electronics conference and exposi-
tion. Austin (TX), USA; March 6–10, 2005. p. 428–34.
[70] Patidar RD, Singh SP, Khatod DK. Single-phase single-stage grid-interactive
photovoltaic system with active filter functions. In: Proceedings of the IEEE
power and energy society general meeting. Minneapolis (MN), USA; 25–29
July, 2010. p. 1–7.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270268

[71] Hirachi K, Mii T, Nakashiba T, Laknath KGD, Nakaoka M. Utility-interactive
multi-functional bidirectional converter for solar photovoltaic power condi-
tioner with energy storage batteries. In: Proceedings of the IEEE industrial
electronics, control, and instrumentation. Taiwan, China; August 5–10, 1996.
p. 1693–8.
[72] Dasgupta S, Sahoo SK, Panda SK. Single-phase inverter control techniques for
interfacing renewable energy sources with microgrid—part I: parallel-
connected inverter topology with active and reactive power flow control
along with grid current shaping. IEEE Transactions on Power Electronics
2011;26(3):717–31.
[73] Chiang SJ, Chang KT, Yen CY. Residential photovoltaic energy storage system.
IEEE Transactions on Industrial Electronics 1998;45(3):385–94.
[74] Bojoi RI, Limongi LR, Roiu D, Tenconi A. Enhanced power quality control
strategy for single-phase inverters in distributed generation systems. IEEE
Transactions on Power Electronics 2011;26(3):798–806.
[75] Limongi LR, Bojoi R, Tenconi A, Clotea L. Single-phase inverter with power
quality features for distributed generation systems. In: Proceedings of the
IEEE conference on optimization of electrical and electronic equipment.
Brasov; May 22–24, 2008. p. 313–8.
[76] Cirrincione M, Pucci M, Vitale G. A single-phase DG generation unit with
shunt active power filter capability by adaptive neural filtering. In: Proceed-
ings of the IEEE industrial electronics conference. Paris, France; 6–10
November, 2006. p. 4373–80.
[77] Macken KJP, Vanthournout K, Van den Keybus J, Deconinck G, Belmans RJM.
Distributed control of renewable generation units with integrated active
filter. IEEE Transactions on Power Electronics 2004;19(5):1353–60.
[78] Hosseini SH, Danyali S, Goharrizi AY. Single stage single phase series-grid
connected PV system for voltage compensation and power supply. In:
Proceedings of the IEEE power and energy society general meeting. Calgary,
Alberta, Canada; 26–30 July, 20 09. p. 1–7.
[79] Mastromauro RA, Liserre M, Dell'Aquila A. PV system power quality
enhancement by means of a voltage controlled shunt-converter. In: Proceed-
ings of the IEEE power electronics specialists conference. Rhodes, Greece;
15–19 June, 2008. p. 2358–63.
[80] Mastromauro RA, Liserre M, Dell'Aquila A. Single-phase grid-connected
photovoltaic systems with power quality conditioner functionality. In:
Proceedings of the IEEE European conference on power electronics and
applications. Aalborg, Denmark; 2–5 September, 2007. p. 1–11.
[81] Vasquez JC, Mastromauro RA, Guerrero JM, Liserre M. Voltage support
provided by a droop-controlled multifunctional inverter. IEEE Transactions
on Industrial Electronics 2009;56(11):4510–9.
[82] Dasgupta S, Sahoo SK, Panda SK, Amaratunga GAJ. Single-phase inverter-
control techniques for interfacing renewable energy sources with microgrid
—part II: series-connected inverter topology to mitigate voltage-related
problems along with active power flow control. IEEE Transactions on Power
Electronics 2011;26(3):732–46.
[83] Lin BR, Yang TY. Implementation of active power filter with asymmetrical
inverter legs for harmonic and reactive power compensation. Electric Power
Systems Research 2005;73(2):227–37.
[84] Geibel D, Degner T, Hardt C, Antchev M, Krusteva A. Improvement of power
quality and reliability with multifunctional PV-inverters in distributed
energy systems. In: Proceedings of the IEEE international conference on
electrical power quality and utilisation. Lodz, Poland; 15–17 September,
2009. p. 1–6.
[85] Kuo YC, Liang TJ, Chen JF. A high-efficiency single-phase three-wire photo-
voltaic energy conversion system. IEEE Transactions on Industrial Electronics
2003;50(1):116–22.
[86] De Souza KCA, Dos Santos WM, Martins DC. A single-phase grid-connected
PV system with active power filter. International Journal of Circuits, Systems
and Signal 2008;2(1):50–5.
[87] De Souza KCA, Dos Santos WM, Martins DC. Optimization of the ferrite core
volume in a single-phase grid-connected PV system with active and reactive
power control. In: Proceedings of the conference on IEEE industrial electro-
nics society. Arizona, USA; 7–10 November, 2010. p. 2803–10.
[88] De Souza KCA, Martins DC. A single-phase active power filter based in a two
stages grid-connected PV system. In: Proceedings of the IEEE power
electronics and motion control conference. Poznan, Poland; 1–3 September,
2008. p. 1951–6.
[89] Wu TF, Shen CL, Chang CH, Chiu J. 1Φ3W grid-connection PV power inverter
with partial active power filter. IEEE Transactions on Aerospace and
Electronic Systems 2003;39(2):635–46.
[90] Wu TF, Shen CL, Nei HS. A 1Ф3W grid-connection PV power inverter with
APF based on nonlinear programming and FZPD algorithm. In: Proceedings
of the applied power electronics conference and exposition. Miami Beach,
Florida, USA; February 9-–13, 2003. p. 546–52.
[91] He JW, Munir MS, Li YW. Opportunities for power quality improvement
through DG-grid interfacing converters. In: Proceedings of the IEEE power
electronics conference. Sapporo, Japan; June 21–24 2010. p. 1657–64.
[92] He JW, Li YW, Munir MS. A flexible harmonic control approach through
voltage-controlled DG —grid interfacing converters. IEEE Transactions on
Industrial Electronics 2012;59(1):444–55.
[93] Yu H, Pan J, Xiang A. A multi-function grid-connected PV system with
reactive power compensation for the grid. Solar Energy 2005;79(1):101–6.
[94] Kim S, Gwonjong Y, Jinsoo S. A bifunctional utility connected photovoltaic
system with power factor correction and UPS facility. In: Proceedings of the
IEEE photovoltaic specialists conference. Washington, DC, USA; 13–17 May ,
1996. p. 1363–8.
[95] Dasgupta S, Mohan SN, Sahoo SK, Panda SK. A Lyapunov function based
current controller to control active and reactive power flow in a three phase
grid connected PV inverter under generalized grid voltage conditions. In:
Proceedings of the IEEE international conference on power electronics and
ECCE Asia. Jeju, Korea; May 30–June 3, 2011. p. 1110–7.
[96] Dasgupta S, Mohan SN, Sahoo SK, Panda SK. Derivation of instantaneous
current references for three phase PV inverter connected to grid with active
and reactive power flow control. In: Proceedings of the IEEE international
conference on power electronics and ECCE Asia. Jeju, Korea; May 30–June 3,
2011. p. 1228–35.
[97] Cheng L, Cheung R, Leung KH. Advanced photovoltaic inverter with addi-
tional active power line conditioning capability. In: Proceedings of the IEEE
power electronics specialists conference. St. Louis, Missouri; June 22–27,
1997. p. 279–83.
[98] Naderi S, Pouresmaeil E, Gao WD. The frequency-independent control
method for distributed generation systems. Applied Energy, in press.
[99] Mohod SW, Aware MV. Micro wind power generator with battery energy
storage for critical load. IEEE Systems Journal 2012;6(1):118–25.
[100] Marei MI, El-Saadany EF, Salama MMA. A novel control algorithm for the DG
interface to mitigate power quality problems. IEEE Transactions on Power
Delivery 2004;19(3):1384–92.
[101] Marei MI, El-Saadany EF, Salama MMA. A flexible DG interface based on a
new RLS algorithm for power quality improvement. IEEE Systems Journal
2011;6(1):68–75.
[102] Cheng PT, Chen CA, Lee TL, Kuo SY. A cooperative imbalance compensation
method for distributed-generation interface converters. IEEE Transactions on
Industry Applications 2009;45(2):805–15.
[103] Lv Z, Luo A, Wu C, Shuai Z, Kang Z. A research of microgrid energy supply and
filtering system based on inverter multiplexing. In: Proceedings of the IEEE
international conference on sustainable power generation and supply.
Nanjing, China; 6–7 April, 2009. p. 1–7.
[104] Luo S, Luo A, Lv Z, Shen Y, Guo L, Jiang W. Power quality active control
research of building integrated photovoltaic. In Proceedings of the IEEE
power electronics for distributed generation systems. Hefei, China; June 16–
18, 2010. p. 796–801.
[105] Mohamed ARIM, El Saadany EF. A control scheme for PWM voltage-source
distributed-generation inverters for fast load-voltage regulation and effective
mitigation of unbalanced voltage disturbances. IEEE Transactions on Indus-
trial Electronics 2008;55(5):2072–84.
[106] Marei MI, Abdel Galil TK, El Saadany EF, Salama MMA. Hilbert transform
based control algorithm of the DG interface for voltage flicker mitigation.
IEEE Transactions on Power Delivery 2005;20(2):1129–33.
[107] Saitou M, Shimizu T. Generalized theory of instantaneous active and reactive
powers in single-phase circuits based on Hilbert transform. In: Proceedings
of the IEEE annual power electronics specialists conference. Cairns, Australia;
23–27 June, 2002. p. 1419–24.
[108] Chandhaket S, Yoshida M, Eiji H, Nakamura M, Konishi Y, Nakaoka M. Multi-
functional digitally-controlled bidirectional interactive three-phase soft-
switching PWM converter with resonant snubbers. In: Proceedings of the
IEEE annual power electronics specialists conference. Vancouver, BC, Canada;
June 17–21, 2001. p. 589–93.
[109] Yamamoto H, Inaba CY, Hiraki E, Nakaoka M. Multifunctional digitally-
controlled bidirectional three-phase soft-switching PWM converter. In:
Proceedings of the international conference on power electronics and
variable speed drives. London, UK; 18–19 September, 2000. p. 86–90.
[110] Prodanovic M, De Brabandere K, Van Den Keybus J, Green T, Driesen J.
Harmonic and reactive power compensation as ancillary services in inverter-
based distributed generation. IET Generation, Transmission and Distribution
2007;1(3):432–8.
[111] Pogaku N, Green TC. Harmonic mitigation throughout a distribution system:
a distributed-generator-based solution. IEE Proceedings: Generation, Trans-
mission and Distribution 2006;153(3):350–8.
[112] Abolhassani MT, Enjeti P, Toliyat H. Integrated doubly fed electric alternator
active filter (IDEA), a viable power quality solution, for wind energy
conversion systems. IEEE Transactions on Energy Conversion 2008;23
(2):642–50.
[113] Todeschini G, Emanuel AE. Transient response of a wind Energy conversion
system used as active filter. IEEE Transactions on Energy Conversion 2011;26
(2):522–31.
[114] Chen Z. Compensation schemes for a SCR converter in variable speed wind
power systems. IEEE Transactions on Power Delivery 2004;19(2):813–21.
[115] Gajanayake CJ, Vilathgamuwa DM, Poh CL, Teodorescu R, Blaabjerg F. Z-
source-inverter-based flexible distributed generation system solution for grid
power quality improvement. IEEE Transactions on Energy Conversion
2009;24(3):695–704.
[116] Tsengenes G, Adamidis G. A multi-function grid connected PV system with
three level NPC inverter and voltage oriented control. Solar Energy 2011;85
(11):2595–610.
[117] Sawant RR, Chandorkar MC. A multifunctional four-leg grid-connected
compensator. IEEE Transactions on Industry Applications 2009;45(1):249–59.
[118] Sawant RR, Chandorkar MC. Methods for multi-functional converter control
in three-phase four-wire systems. IET Power Electronics 2009;2(1):52–66.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270 269

[119] Wang F, Duarte JL, Hendrix M. Control of grid-interfacing inverters with
integrated voltage unbalance correction. In: Proceedings of the IEEE power
electronics specialists conference. Rodes, Greece; 15–19 June 2008. p. 310–6.
[120] Wang F, Duarte JL, Hendrix MAM. Pliant active and reactive power control for
grid-interactive converters under unbalanced voltage dips. IEEE Transactions
on Power Electronics 2011;26(5):1511–21.
[121] Pinto JG, Pregitzer R, Monteiro LFC, Afonso JL. 3-Phase 4-wire shunt active
power filter with renewable energy interface. In: Proceedings of interna-
tional conference on renewable energies and power quality. Seville, Spain;
28–30 March, 20 07. p. 1–6.
[122] Majumder R, Ghosh A, Ledwich G, Zare F. Load sharing and power quality
enhanced operation of a distributed microgrid. IET Renewable Power Gen-
eration 2009;3(2):109–19.
[123] Shahnia F, Majumder R, Ghosh A, Ledwich G, Zare F. Operation and control of
a hybrid microgrid containing unbalanced and nonlinear loads. Electric
Power Systems Research 2010;80(8):954–65.
[124] Han B, Bae B, Kim H, Baek S. Combined operation of unified power-quality
conditioner with distributed generation. IEEE Transactions on Power Deliv-
ery 2006;21(1):330–8.
[125] Li YW, Vilathgamuwa DM, Loh PC. Microgrid power quality enhancement
using a three-phase four-wire grid-interfacing compensator. IEEE Transac-
tions on Industry Applications 2005;41(6):1707–19.
[126] Li YW, Vilathgamuwa DM, Loh PC. A grid-interfacing power quality com-
pensator for three-phase three-wire microgrid applications. IEEE Transac-
tions on Power Electronics 2006;21(4):1021–31.
[127] Wang F, Duarte JL, Hendrix MAM. Grid-interfacing converter systems with
enhanced voltage quality for microgrid application—concept and implemen-
tation. IEEE Transactions on Power Electronics 2011;26(12):3501–13.
[128] Shen G, Xu D, Cao L, Zhu X. An improved control strategy for grid-connected
voltage source inverters with an LCL filter. IEEE Transactions on Power
Electronics 2008;23(4):1899–906.
[129] Yu XX, Khambadkone AM. Multi-functional power CONVERTER building
block to facilitate the connection of micro-grid. In: Proceedings of the IEEE
workshop on control and modeling for power electronics. Zurich, Switzer-
land; 17–20 August, 2008. p. 1–6.
[130] Zhou K, Wang D. Relationship between space-vector modulation and three-
phase carrier-based PWM: a comprehensive analysis. IEEE Transactions on
Industrial Electronics 2002;49(1):186–96.
[131] Zeng QR, Chang LC. An advanced SVPWM-based predictive current controller
for three-phase inverters in distributed generation systems. IEEE Transac-
tions on Industrial Electronics 2008;55(3):1235–46.
[132] Kang BJ, Liaw CM. Robust hysteresis current-controlled PWM scheme with
fixed switching frequency. IEE Proceedings—Electric Power Applications
2001;148(6):503–12.
[133] Dannehl J, Fuchs FW, Th XF, Gersen PB. PI state space current control of grid-
connected PWM converters with LCL filters. IEEE Transactions on Power
Electronics 2010;25(9):2320–30.
[134] Zmood DN, Holmes DG. Stationary frame current regulation of PWM
inverters with zero steady-state error. IEEE Transactions on Power Electro-
nics 2003;18(3):814–22.
[135] Qi C, Yang Y, Suo J. Deadbeat decoupling control of three-phase photovoltaic
grid-connected inverters. In: Proceedings of the IEEE international confer-
ence on mechatronics and automation. Changchun, Jilin, China; 9–12 August,
2009. p. 3843–8.
[136] Moreno JC, Huerta JME, Gil RG, Gonzalez SA. A robust predictive current
control for three-phase grid-connected inverters. IEEE Transactions on
Industrial Electronics 2009;56(6):1993–2004.
[137] Miranda H, Teodorescu R, Rodriguez P, Helle L. Model predictive current
control for high-power grid-connected converters with output LCL filter. In:
Proceedings of the annual conference of IEEE on industrial electronics. Porto,
Portugal; 3–5 November, 2009. p. 633–8.
[138] Dasgupta S, Sahoo SK, Panda SK. Design of a spatial iterative learning
controller for single phase series connected PV module inverter for grid
voltage compensation. In: Proceedings of the IEEE international power
electronics conference. Sapporo, Japan; 21–24 June, 2010. p. 1980–7.
[139] Hornik T, Zhong QC. H∞repetitive voltage control of grid-connected
inverters with a frequency adaptive mechanism,. IET Power Electronics
2010;3(6):925–35.
[140] Guo X, Wu W, Gu H. Modeling and simulation of direct output current
control for LCL-interfaced grid-connected inverters with parallel passive
damping. Simulation Modelling Practice and Theory 2010;18(7):946–56.
[141] Asiminoael L, Blaabjerg F, Hansen S. Detection is key—harmonic detection
methods for active power filter applications. IEEE Industry Applications
Magazine 2007;13(4):22–33.
[142] Rechka S, Ngandui E, Xu J, Sicard P. Performance evaluation of harmonics
detection methods applied to harmonics compensation in presence of
common power quality problems. Mathematics and Computers in Simulation
2003;63(3–5):363–75.
[143] Montero MIM, Cadaval ER, Gonzalez FB. Comparison of control strategies for
shunt active power filters in three-phase four-wire systems. IEEE Transac-
tions on Power Electronics 2007;22(1):229–36.
[144] Solomon OM. The use of DFT windows in signal-to-noise ratio and harmonic
distortion computations. IEEE Transactions on Instrumentation and Mea-
surement 1994;43(2):194–9.
[145] Morsi WG, El Hawary ME. Defining power components in nonsinusoidal
unbalanced polyphase systems: the issues. IEEE Transactions on Power
Delivery 2007;22(4):2428–38.
[146] Depenbrock M. The FBD-method, a generally applicable tool for analyzing
power relations. IEEE Transactions on Power Systems 1993;8(2):381–7.
[147] Akagi H, Watanabe EH, Aredes M. Instantaneous power theory and applica-
tions to power conditioning. Hoboken, New Jersey: John Wiley and Sons, Inc.;
2007.
[148] Furuhashi T, Okuma S, Uchikawa Y. A study on the theory of instantaneous
reactive power. IEEE Transactions on Industrial Electronics 1990;37(1):86–90.
[149] Srianthumrong S, Fujita H, Akagi H. Stability analysis of a series active filter
integrated with a double-series diode rectifier. IEEE Transactions on Power
Electronics 2002;17(1):117–24.
[150] Pereira RR, Da Silva CH, Da Silva LEB, Lambert-Torres G, Pinto JOP. New
strategies for application of adaptive filters in active power filters. IEEE
Transactions on Industry Applications 2011;47(3):1136–41.
[151] Rechka S, Ngandui E, Xu J, Sicard P. Analysis of harmonic detection
algorithms and their application to active power filters for harmonics
compensation and resonance damping. Canadian Journal of Electrical and
Computer Engineering 2003;28(1):41–51.
[152] Han Y, Mansoor G, Yao L, Zhou, Chen C. Design and experimental investiga-
tion of a three-phase APF based on feed-forward plus feedback control.
WSEAS Transactions on Power Systems 2008;2(3):15–20.
Z. Zeng et al. / Renewable and Sustainable Energy Reviews 24 (2013) 223–270270