Benchmarking of Constant Power Generation Strategies for Single-Phase Grid-Connected Photovoltaic Systems

Abstract

With a still increase of grid-connected Photovoltaic (PV) systems, challenges have been imposed on the grid due to the continuous injection of a large amount of fluctuating PV power, like overloading the grid infrastructure (e.g., transformers) during peak power production periods. Hence, advanced active power control methods are required. As a cost-effective solution to avoid overloading, a Constant Power Generation (CPG) control scheme by limiting the feed-in power has been introduced into the currently active grid regulations. In order to achieve a CPG operation, this paper presents three CPG strategies based on: 1) a power control method (P-CPG), 2) a current limit method (I-CPG) and 3) the Perturb and Observe algorithm (P&O-CPG). However, the operational mode changes (e.g., from the maximum power point tracking to a CPG operation) will affect the entire system performance. Thus, a benchmarking of the presented CPG strategies is also conducted on a 3-kW single-phase grid-connected PV system. Comparisons reveal that either the P-CPG or I-CPG strategies can achieve fast dynamics and satisfactory steady-state performance. In contrast, the P&O-CPG algorithm is the most suitable solution in terms of high robustness, but it presents poor dynamic performance.

Full text

Aalborg Universitet
Benchmarking of Constant Power Generation Strategies for Single-Phase Grid-
Connected Photovoltaic Systems
Sangwongwanich, Ariya; Yang, Yongheng; Blaabjerg, Frede; Wang, Huai
Published in:
I E E E Transactions on Industry Applications
Publication date:
2018
Document Version
Accepted author manuscript, peer reviewed version
Link to publication from Aalborg University
Citation for published version (APA):
Sangwongwanich, A., Yang, Y., Blaabjerg, F., & Wang, H. (2018). Benchmarking of Constant Power Generation
Strategies for Single-Phase Grid-Connected Photovoltaic Systems. I E E E Transactions on Industry
Applications, PP(99), 1-11.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
? Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
? You may not further distribute the material or use it for any profit-making activity or commercial gain
? You may freely distribute the URL identifying the publication in the public portal ?
Take down policy
If you believe that this document breaches copyright please contact us at vbn@aub.aau.dk providing details, and we will remove access to
the work immediately and investigate your claim.
Downloaded from vbn.aau.dk on: august 21, 2017

IEEE INDUSTRY APPLICATIONS MAGAZINE, VOL. PP, NO. 99, 2016 1
Benchmarking of Constant Power Generation Strategies for
Single-Phase Grid-Connected Photovoltaic Systems
Ariya Sangwongwanich, Student Member, IEEE, Yongheng Yang, Member, IEEE, Frede Blaabjerg, Fellow, IEEE
and Huai Wang, Member, IEEE
Abstract—With a still increase of grid-connected Photovoltaic
(PV) systems, challenges have been imposed on the grid due to the
continuous injection of a large amount of fluctuating PV power,
like overloading the grid infrastructure (e.g., transformers)
during peak power production periods. Hence, advanced active
power control methods are required. As a cost-effective solution
to avoid overloading, a Constant Power Generation (CPG) control
scheme by limiting the feed-in power has been introduced into
the currently active grid regulations. In order to achieve a CPG
operation, this paper presents three CPG strategies based on:
1) a power control method (P-CPG), 2) a current limit method
(I-CPG) and 3) the Perturb and Observe algorithm (P&O-CPG).
However, the operational mode changes (e.g., from the maximum
power point tracking to a CPG operation) will affect the entire
system performance. Thus, a benchmarking of the presented
CPG strategies is also conducted on a 3-kW single-phase grid-
connected PV system. Comparisons reveal that either the P-CPG
or I-CPG strategies can achieve fast dynamics and satisfactory
steady-state performance. In contrast, the P&O-CPG algorithm
is the most suitable solution in terms of high robustness, but it
presents poor dynamic performance.
Index Terms—Active power control, constant power control,
maximum power point tracking, PV systems, power converters.
I. INTRODUCTION
Photovoltaic (PV) systems have a high growth rate during
the last several years, and will play an even more significant
role in the future mixed power grid [1]–[3]. A majority of PV
system is connected to the distribution grid (i.e., mainly single-
phase systems) [2] where a Maximum Power Point Tracking
(MPPT) is currently mandatory in most active grid codes, and
also to ensure the maximum energy yield from the solar power
[4]. At a high penetration level of PV systems in the near
future, the grid may face a challenge of overloading during
peak power generation periods through a day if the power
capacity of the grid remains the same [5]–[7]. For instance, it
was reported by BBC that parts of the Northern Ireland’s grid
Manuscript received April 15, 2016; revised June 29, 2016; accepted
September 28, 2016. Paper 2016-SECSC-0338.R1, presented at the 2016
IEEE Applied Power Electronics Conference and Exposition, Long Beach,
California, March 20-24, 2016, and approved for publication in the IEE E
INDUS TRY APP LICATIO NS MAGAZI NE by the Sustainable Energy
Conversion Systems Committee of the IEEE Industry Applications Society.
This work was supported in part by the European Commission within the
European Unions Seventh Frame-work Program (FP7/2007-2013) through
the SOLAR-ERA.NET Transnational Project (PV2.3 - PV2GRID), by En-
erginet.dk (ForskEL, Denmark, Project No. 2015-1-12359), and in part by the
Research Promotion Foundation (RPF, Cyprus, Project No. KOINA/SOLAR-
ERA.NET/0114/02).
The authors are with the Department of Energy Technology, Aalborg Uni-
versity, DK-9220 Aalborg, Denmark (e-mail: ars@et.aau.dk; yoy@et.aau.dk;
fbl@et.aau.dk; hwa@et.aau.dk).
This is the reference copy of the accepted version. When it is published,
color versions of one or more of the figures in this paper will be available
online at http://ieeexplore.ieee.org.
Fig. 1. Constant Power Generation (CPG) concept for PV systems: 1) MPPT
mode during I, III, V, and 2) CPG mode during II, IV [15].
were overloaded by the increased number of grid-connected
PV systems in a sunny and clear day with strong solar
irradiance [8]. In order to enable more PV installations and
address such issues, the control algorithms have to be feasible
to flexibly regulate the active power generated by PV systems
[4], [9]–[12]. For instance, limiting the feed-in power of PV
systems to a certain level has been found as an effective
approach to overcome overloading [10], and thus it is currently
required in Germany through the grid codes [13], where it is
stated that newly installed PV systems with a rated power
below 30 kWp have to be able to limit its maximum feed-
in power (i.e., 70% of the rated power) unless it can be
remotely controlled. In fact, this active power control strategy
corresponds to an absolute power constraint defined in the
Danish grid code [14], and it is also referred to as a Constant
Power Generation (CPG) control in the prior-art work [15].
Actually, there are several methods to limit the feed-in
power of the PV system in order to achieve a constant power
production (e.g., integrating energy storage systems, installing
dump load) [16]. However, the most intuitive and cost-effective
way to achieve the CPG control is through the modification
of the MPPT algorithm at the PV inverter level (also called
power curtailment), and will be considered in this paper [17].
In this approach, the PV system continues operating in the
MPPT mode with injection of the maximum power as long
as the available PV power PMPPT is below the set-point
Plimit. However, when the available power reaches the level
of Plimit, the PV system will inject a constant active power,
i.e., Ppv =Plimit. The operational principle of the CPG scheme
can be illustrated in Fig. 1 and summarized as:
Ppv =PMPPT,when PMPPT ≤Plimit
Plimit,when PMPPT > Plimit
(1)
where Ppv is the PV output power, PMPPT is the maximum
available power (according to the MPPT operation), and Plimit

2 IEEE INDUSTRY APPLICATIONS MAGAZINE, VOL. PP, NO. 99, 2016
Fig. 2. Possible operating points of the PV system in Power-Voltage curve
of the PV arrays during the CPG operation (i.e., Constant Power Point) at a
certain level of power limit Plimit and irradiance.
Fig. 3. Hardware schematics and overall control structure of a two-stage
single-phase grid-connected PV system.
is the power limit, which is the set-point. The constant power
production can be achieved by regulating the PV output
power at the operating point below the Maximum Power Point
(MPP), as it is shown in Fig. 2, and this operating point is
called the Constant Power Point (CPP) in this paper [18].
In the prior-art work, several CPG strategies for PV systems
have been introduced. In fact, more methods have also been
proposed for other applications (e.g., frequency regulation,
low-voltage ride through), but they can also be applied to
achieve the CPG control as well. Accordingly, the CPG
strategies presented in literature can be generally classified
into three different approaches. In [15], [19]–[22], the CPG
control is realized by directly regulating the PV power to be
constant through the closed-loop power control. This can be
implemented either at the dc-dc stage [15], [19], [22], where
the boost converter is controlled directly, or at the dc-ac stage
[20], [21], where a constant power reference Plimit is applied
to the P Q controller of the PV inverter. Another way to limit
the power generated from the PV systems is by controlling the
PV output current ipv, as it is discussed in [23] and [24]. This
approach is based on the characteristic of the PV arrays where
the PV output current ipv is strongly dependent on the solar
irradiance level, while the PV output voltage vpv varies only in
a small range during irradiance change. Thus, limiting the PV
output current ipv can effectively limit the PV output power
Ppv. Alternatively, the CPG operation can also be realized
by using the Perturb and Observe (P&O) algorithm, as it is
proposed in [25]–[28]. In this method, the PV output voltage
vpv is continuously perturbed away from the MPP during the
CPG operation mode, in order to reduce the PV output power
according to the set-point (i.e., Ppv =Plimit).
TABLE I
PARA MET ER S OF TH E TWO -STAG E SINGLE-PH AS E PV SY ST EM (F IG . 3).
PV rated power 3 kW
Boost converter inductor L= 1.8 mH
PV-side capacitor Cpv = 1000 µF
DC-link capacitor Cdc = 1100 µF
LCL-filter Linv = 4.8 mH, Lg= 2 mH,
Cf= 4.3 µF
Switching frequency Boost converter: fb= 16 kHz,
Full-Bridge inverter: finv = 8 kHz
DC-link voltage v∗
dc = 450 V
Grid nominal voltage (RMS) Vg= 230 V
Grid nominal frequency ω0= 2π×50 rad/s
Nevertheless, the performance of the three CPG approaches
have not yet been compared. Thus, it is difficult to justify
which method is suitable to be implemented in industry and
applied in the future grid codes. Besides, most of the literatures
only discuss the performance of the CPG strategy during
steady-state (e.g., during a constant irradiance condition). In
fact, depending on the mission profiles of the PV system (e.g.,
irradiance and temperature conditions), the operation mode
transition between the MPPT and CPG can challenge the
system performance, especially during a fluctuating irradiance
condition (e.g., in a cloudy day). This will affect the system
performance in terms of dynamics, accuracy, and stability of
the CPG strategy, which have not yet been investigated so far.
In the light of the above issues, this paper first discusses
about three different CPG strategies applied to two-stage
single-phase PV systems. Then, the performance of the CPG
strategies under both dynamic and steady-state conditions are
benchmarked experimentally on a 3-kW two-stage single-
phase grid-connected PV system, where real-field mission
profiles are taken into consideration. Finally, conclusions are
drawn from the comparison in §V.
II. CO NT ROL ST RUCTURE OF TWO -STAG E SINGLE-PHA SE
GRID-CON NE CT ED PV SY ST EMS
A. System Configuration
In most single-phase PV systems (e.g., rated power of 1 -
30 kW), a two-stage configuration is widely used [29], [30].
The system configuration and its control structure are shown in
Fig. 3, where the system parameters are given in Table I. The
PV arrays are connected to a boost converter, allowing a wide-
range operation during both MPPT and CPG operations [31].
In other words, with the use of the two-stage configuration,
the PV system can operate at a lower PV voltage vpv (e.g., at
the left side of the MPP in the case of the CPG operation),
since the PV output voltage vpv can be stepped up by the
boost converter to match the required dc-link voltage (e.g.,
450 V) for the PV inverter [29]. This may not be possible in
the single-stage configuration, where the PV output voltage
vpv is directly fed to the PV inverter (i.e., vpv =vdc with vdc
being the dc-link voltage). Practically, the dc-link voltage vdc
is required to be higher than the peak grid voltage level (e.g.,
325 V) to ensure the power delivery [32].

SANGWONGWANICH et al.: BENCHMARKING OF CPG STRATEGIES FOR SINGLE-PHASE GRID-CONNECTED PV SYSTEMS 3
Fig. 4. Implementation of different MPPT controllers: (a) PV output voltage,
(b) PV output current, and (c) PV output power, where PI represents a
proportional-integral controller.
In the boost converter stage, either the MPPT or CPG
control can be implemented in order to control the power
extraction from the PV arrays. Then, the extracted power is
delivered to the ac grid through the control of the full-bridge
inverter. In this case, the control of the full-bridge inverter
keeps the dc-link voltage to be constant through the control
of the injected grid current [33].
B. Boost Converter Controller
As aforementioned, the boost converter plays a major role to
control the power extraction from the PV arrays. Therefore, it
is important to discuss about the possible control structures
for the boost converter, where the CPG strategies will be
implemented. Usually, the MPPT algorithm (i.e., the P&O
MPPT) is implemented in the boost converter. For example,
the P&O MPPT algorithm can give either the reference PV
voltage vMPPT or current iMPPT to control the boost converter.
Thus, the MPPT is usually achieved by regulating either the
PV output voltage vpv or current ipv according to the reference
from the MPPT algorithm, as it is shown in Figs. 4(a) and (b).
Alternatively, it is possible to achieve the MPPT through the
control of the PV output power Ppv. In this case, the reference
PV current from the MPPT algorithm iMPPT is multiplied by
the measured PV voltage vpv in order to obtained the reference
PV power PMPPT, as it is shown in Fig. 4(c). In this way, the
PV power Ppv is controlled directly at any time, making it
possible and flexible to be modified according to the power
set-point (e.g., to realize the CPG operation). However, it
should be mentioned that the variations in the PV voltage
vpv (e.g., due to the noise from measurements) can propagate
to the reference PMPPT through the direct multiplication, and
thereby decrease the tracking accuracy of the MPPT operation.
Nevertheless, the tracking errors are with permissible limits,
which will be experimentally verified in §IV.
It is noteworthy to mention that tracking the MPP by
controlling the PV current ipv (i.e., Fig. 4(b)) is of less
robustness [34]. This is due to the very steep slope (i.e., large
Fig. 5. Stability issues of the MPPT controller based on the PV output current
due to the high slope (dPpv/dipv ) at the right side of the MPP [34].
dPpv/dipv ) on the right side of the MPP in the power-current
(P−I) curve of the PV arrays, as it is shown in Fig. 5. The
operating point of the PV system may go into the short-circuit
condition under a sudden decrease of the irradiance condition
(if the MPPT algorithm cannot track fast enough, e.g., the PV
system still operating at the same PV output current ipv), when
the PV output current is controlled [34]. This can be illustrated
by the A→B trajectory in Fig. 5 when the irradiance level
suddenly drops from 1000 W/m2to 700 W/m2. This stability
issue will be observed in a CPG control scheme which is based
on the control structure in Fig. 4(b).
III. CON STAN T POW ER GENERATION STRATE GI ES
Basically, the CPG strategy needs to regulate the operating
point of the PV system at the CPPs in order to achieve a con-
stant power production. According to the P−Vcharacteristic
curve of the PV arrays shown in Fig. 2, there are two possible
operating points –CPP-L and CPP-R for the CPG mode at a
certain power level (i.e., Plimit) and a certain irradiance level.
However, the CPPs continuously change (i.e., different PV
voltage and PV current) under a changing irradiance condition,
according to the P−Vcurve of the PV arrays. Thus, the CPG
strategy has to be able to follow the change in the P−Vcurve,
and track the CPP in the case of the CPG operation. Generally,
the demands for the CPG control schemes are
•In the steady-state CPG operation, the CPG strategies
should keep the PV systems operating at one of the CPPs
with minimum deviations, in order to minimize the power
losses in the steady-state.
•Under a changing irradiance condition (e.g., in a cloudy
day), the CPG control scheme should be able to track
either the MPP or the CPP, depending on the operating
mode, and at the same time ensure a stable transition.
Accordingly, three previously mentioned CPG strategies are
adapted to two-stage single-phase PV systems, and are dis-
cussed in the following based on: 1) a power control method
(P-CPG), 2) a current limit method (I-CPG), and 3) the Perturb
and Observe algorithm (P&O-CPG), where the above demands
are taken as the benchmarking criteria.
A. CPG based on a Power Control Method (P-CPG)
Limiting the PV output power through the closed-loop
power control is one of the most commonly used solutions

4 IEEE INDUSTRY APPLICATIONS MAGAZINE, VOL. PP, NO. 99, 2016
Fig. 6. Control structure of the Constant Power Generation (CPG) scheme
based on a power control (P-CPG).
to achieve the CPG control in the previous work [15], [19].
In order to realize this control method in the two-stage
single-phase PV system, the boost converter needs to directly
control the PV output power during operation. As mentioned
previously, it is possible to directly control the PV output
power Ppv during the MPPT operating mode by employing
the control scheme in Fig. 4(c), where the reference PV power
in the MPPT mode PMPPT is obtained by multiplying the
reference current iMPPT from the MPPT algorithm with the
PV voltage vpv. Regarding the CPG operation, a saturation
block is added to the control scheme in Fig. 4(c) in order
to limit the reference PV power P∗
pv to a certain power level
Plimit, as it is shown in Fig. 6. Namely, when the reference
PV power from the MPPT algorithm PMPPT reaches the level
of power limit Plimit, the saturation block will keep the power
reference to be constant, i.e., P∗
pv =Plimit, and the PV system
enters into the CPG mode. Otherwise, if the the reference PV
power from the MPPT algorithm PMPPT is less than the power
limit Plimit, the saturation block will not be activated, and the
PV system will operate in the MPPT mode with a maximum
power injection (i.e., P∗
pv =PMPPT), which is equivalent to the
MPPT controller in Fig. 4(c). The operational principle can be
further summarized as:
P∗
pv =PMPPT,when PMPPT ≤Plimit
Plimit,when PMPPT > Plimit
(2)
where PMPPT is the maximum available power (according to
the MPPT operation), and Plimit is the power limit, as defined
previously.
B. CPG based on a Current Limit Method (I-CPG)
Another way to control the PV output power is through the
control of the PV output current ipv as it is discussed in [23],
[24]. This is due to the fact that the PV voltage vpv only varies
in a small range during the irradiance change in the operating
region on right side of the MPP (at the CPP-R), as it is shown
in Fig. 7. Therefore, the PV output power Ppv can effectively
be controlled through the PV output current ipv in this region.
From the control scheme in Fig. 4(b), it is possible to achieve
a CPG operation by limiting the reference current from the
MPPT algorithm iMPPT according to ilimit =Plimit/vpv when
calculating the reference PV output current i∗
pv. The control
structure of the I-CPG method is shown in Fig. 8, and the
power limit Plimit corresponds to the rectangular area under
the CPP-R in Fig. 7.
According to the CPG concept in (1), the performance
of the controller during the MPPT operation should not be
Fig. 7. Operational principle of the Constant Power Generation (CPG) scheme
based on a current limit (I-CPG).
Fig. 8. Control structure of the Constant Power Generation (CPG) scheme
based on a current limit (I-CPG).
diminished by the current limit. This can be ensured when
considering
PMPPT
vpv
≤Plimit
vpv
and thus,
iMPPT ≤ilimit
where it can be seen that the current limit will not be activated
as long as PMPPT ≤Plimit, and the I-CPG method in the MPPT
mode is simply equivalent to the MPPT controller in Fig. 4(b).
C. CPG based on the P&O Algorithm (P&O-CPG)
A CPG operation can also be realized by means of a Perturb
and Observe (P&O) algorithm [25]–[28]. This method is based
on the MPPT control structure in Fig. 4(a), where the PV
voltage vpv is controlled. In this approach, the modification is
done at the control algorithm when determining the reference
PV voltage v∗
pv. More precisely, during the MPPT operation,
the reference PV voltage v∗
pv is set from the MPPT algorithm
(i.e., P&O MPPT). However, in the case of the CPG operation,
the PV voltage vpv is continuously perturbed towards one CPP,
i.e., Ppv =Plimit, as illustrated in Fig. 9. After a number of
iterations, the operating point will be reached and oscillate
around the corresponding CPP. Notably, the two-stage PV
system with the P&O-CPG control can operate at either the
CPP-L or the CPP-R, depending on the perturbation direction
of the algorithm. However, the power oscillation in the steady-
state is larger at the CPP-R compared to that at the CPP-L
due to the high slope of the P−Vcurve on the right side of
the MPP (i.e., large dPpv/dvpv ). This large power oscillation
will decrease the tracking accuracy, and increase the energy
losses as well as the power fluctuations in the steady-state,
which should be avoided. On the other hand, the operating

SANGWONGWANICH et al.: BENCHMARKING OF CPG STRATEGIES FOR SINGLE-PHASE GRID-CONNECTED PV SYSTEMS 5
Fig. 9. Operational principle of the Constant Power Generation (CPG) scheme
based on the P&O algorithm (P&O-CPG).
Fig. 10. Control structure of the Constant Power Generation (CPG) scheme
based on the P&O algorithm (P&O-CPG).
region at the CPP-L requires a higher conversion ratio (i.e.,
vdc/vpv ) which may affect the boost converter efficiency [35].
The control structure of the algorithm is shown in Fig. 10,
where the reference PV voltage v∗
pv can be expressed as:
v∗
pv =vMPPT,when Ppv ≤Plimit
vpv −vstep,when Ppv > Plimit
(3)
if the PV system operates at the CPP-L, or
v∗
pv =vMPPT,when Ppv ≤Plimit
vpv +vstep,when Ppv > Plimit
(4)
if the PV system operates at the CPP-R, where vMPPT is the
reference voltage from the MPPT algorithm (i.e., the P&O
MPPT algorithm) and vstep is the perturbation step size.
IV. BENCHMARKING OF CON STA NT POW ER GE NE RATI ON
(CPG) STR ATEG IE S
In order to benchmark the discussed CPG control strategies,
experiments have been carried out referring to Fig. 3, where
the experimental test-rig is shown in Fig. 11. The performance
of the two-stage single-phase PV system during the MPPT
operation are demonstrated in Fig. 12(a). Here, the sampling
frequency of the MPPT (and also CPG) algorithms is chosen as
10 Hz (which is a typical sampling rate of the MPPT algorithm
[36]). For the PV inverter controller, the dc-link voltage vdc is
regulated at 450 ±5 V and the extracted power is delivered
to a single-phase 50-Hz ac grid with a peak voltage of 325 V,
as it can be seen from Fig. 12(b).
In the experiments, a 3-kW PV simulator has been adopted,
where irradiance and ambient temperature profiles can be
programmed to emulate the behavior of real PV arrays in
different operating conditions. First, the performance of the
CPG strategies are examined with a slow changing trapezoidal
solar irradiance profile in Fig. 13, where three different values
of power limit Plimit (i.e., 20 %, 50 %, and 80 % of the rated
power) are used to verify the feasibility of the CPG strategies
Fig. 11. Experimental setup of the two-stage grid-connected PV system.
Fig. 12. Performance of the two-stage single-phase grid-connected PV
system: (a) the PV power extraction during the MPPT operation, and (b) the
grid voltage vg, grid current igand the phase angle θduring the steady-state
MPPT operation (3 kW).
under various set-points. Then, a fast changing trapezoidal
solar irradiance profile in Fig. 14 is adopted, in order to
challenge the dynamic of the CPG strategy and to observe the
behavior of the algorithm during the operating mode transition
(e.g., from MPPT to CPG mode). Furthermore, two real-
field solar irradiance and ambient temperature profiles are also
programmed in order to examine the performance of the CPG
algorithms in the real operation, where Plimit = 1.5 kW (i.e.,
50 % of the rated power). A clear day irradiance condition
is used in Fig. 15, where the solar irradiance level changes
relatively slowly and smoothly. In this condition, the CPG
strategy mostly operates in steady-state condition. In contrast,
the dynamic performance of the CPG strategy can clearly be
seen during the fluctuating irradiance condition in Fig. 16,

6 IEEE INDUSTRY APPLICATIONS MAGAZINE, VOL. PP, NO. 99, 2016
Fig. 13. Experimental results of the Constant Power Generation (CPG) scheme based on: (a) the power control, (b) the current limit, (c) the P&O at the
CPP-R, and (d) the P&O at the CPP-L under a slow changing irradiance condition. The tracking error is calculated from the difference between the actual
PV output power Ppv and its set-point Plimit = 80 % during the CPG mode (i.e., |Ppv −Plimit|), and then divided by the total energy yield.
Fig. 14. Experimental results of the Constant Power Generation (CPG) scheme based on: (a) the power control, (b) the current limit, (c) the P&O at the
CPP-R, and (d) the P&O at the CPP-L under a fast changing irradiance condition. The tracking error is calculated from the difference between the actual PV
output power Ppv and its set-point Plimit = 80 % during the CPG mode (i.e., |Ppv −Plimit|), and then divided by the total energy yield.

SANGWONGWANICH et al.: BENCHMARKING OF CPG STRATEGIES FOR SINGLE-PHASE GRID-CONNECTED PV SYSTEMS 7
Fig. 15. Experimental results of the Constant Power Generation (CPG) scheme based on: (a) the power control, (b) the current limit, (c) the P&O at the
CPP-R, and (d) the P&O at the CPP-L under a clear day condition. The tracking error is calculated from the difference between the actual PV output power
Ppv and its set-point Plimit = 1.5 kW during the CPG mode (i.e., |Ppv −Plimit|), and then divided by the total energy yield.
Fig. 16. Experimental results of the Constant Power Generation (CPG) scheme based on: (a) the power control, (b) the current limit, (c) the P&O at the
CPP-R, and (d) the P&O at the CPP-L under a cloudy day condition. The tracking error is calculated from the difference between the actual PV output power
Ppv and its set-point Plimit = 1.5 kW during the stable CPG mode (i.e., |Ppv −Plimit|), and then divided by the total energy yield.

8 IEEE INDUSTRY APPLICATIONS MAGAZINE, VOL. PP, NO. 99, 2016
Fig. 17. Trajectory of the operating point of the Constant Power Generation (CPG) scheme based on: (a) the power control, (b) the current limit, (c) the
P&O at the right side of the MPP, and (d) the P&O at the left side of the MPP under a slow changing irradiance condition (Fig. 13), when Plimit = 2.4 kW.
where the cloudy day irradiance profile is emulated. During
the above tests, the average tracking error (in percentage of
the total energy yield) during the CPG mode is also provided
in the same figure. The tracking error is calculated from
the difference between the actual PV output power and its
set-point (i.e., |Ppv −Plimit|), and then divided by the total
energy yield in order to make it comparable for different test
conditions. This parameter can be used for comparing the
tracking accuracy of different CPG strategies numerically. For
instance, a large value of tracking error indicates a violation
of the CPG constraint (i.e., Ppv > Plimit ) and/or significant
energy losses (i.e., Ppv < Plimit ). Fig. 17 shows an example
of the operating trajectories of the CPG strategies, where the
irradiance condition in Fig. 13 is used. The detailed discussion
about the results are given and benchmarked in the following.
A. Dynamic responses
The dynamic responses can be observed during the CPG
to MPPT transition and vice versa. For the trapezoidal ir-
radiance condition, this transition occurs when the available
power reaches the level of power limit Plimit . In Fig. 13
all the CPG strategies have a smooth transition, since the
irradiance changes relatively slowly. However, in the case
of fast changing solar irradiance in Fig. 14, the dynamics
of the CPG strategies are more challenged to follow the
changes in the CPP. It can be observed from Figs. 14(c) and
(d) that the P&O-CPG scheme presents large power over-
shoots during the MPPT to CPG transition. This is due to the
fact that the P&O-CPG scheme is an iteration-based method,
which requires a number of iterations in order to reach the
corresponding CPP. A long-term dynamic response can be
examined with the cloudy day irradiance condition in Fig. 16,
where PV output power is continuously fluctuating. In this
condition, similar power overshoots also appear in the P&O-
CPG algorithm as it can be seen in Figs. 16(c) and (d). In
contrast, the P- and I-CPG algorithms can regulate the PV
output power to be constant almost without any overshoots
during both short-term (i.e., Fig. 14) and long-term (i.e., Fig.
16) fast changing irradiance conditions. This fast dynamic
performance is achieved because the P- and I-CPG strategies
directly regulate the corresponding reference PV power Ppv
(i.e., P-CPG) or PV current ipv (i.e., I-CPG) through the
close-loop control during the CPG mode. In other words, the
algorithms do not require iterations in order to reach the CPP.
B. Steady-state responses
In the steady-state, the CPG algorithm should regulate the
PV power Ppv to be constant with minimum deviations, as
discussed in §III. This can be observed from Figs. 13 and 14
during the time period when the irradiance level is constant.
A long-term steady state performance can also be seen in Fig.
15, since the irradiance level changes slowly and smoothly
in the clear day condition. The experimental results in the
above conditions show that most of the CPG algorithms have
a satisfactory steady-state performance, where the PV output
power Ppv is limited according to the set-point Plimit with very
small deviations. However, when the P&O-CPG algorithm is
employed to regulate the PV power at the right side of the

SANGWONGWANICH et al.: BENCHMARKING OF CPG STRATEGIES FOR SINGLE-PHASE GRID-CONNECTED PV SYSTEMS 9
MPP (i.e., at the CPP-R), large power oscillations appear as
shown in Figs. 13(c), 14(c), and 15(c). This is due to the large
dPpv/dvpv at the CPP-R (see Fig. 2). Actually, it can be noticed
from Figs. 13(c) and 14(c) that the power oscillation becomes
even larger at the low power limit level (e.g., when Plimit =
20%), as the slope dPpv/dvpv increases when the operating
point is further at the right side of the MPP.
C. Tracking error
The tracking error is another important performance aspect
of the CPG strategy, which indicates numerically how well
the algorithm follows the change in the CPP during the CPG
operation. In fact, the tracking error is a consequence of both
the dynamic and steady-state responses, depending on the
irradiance profile. For example, the tracking error in steady-
state is dominant in the trapezoidal irradiance profiles in Figs.
13 and 14, since the time period of a constant irradiance is
much longer than the ramp-changing (considered only during
the CPG mode). Therefore, the tracking errors of the P&O-
CPG strategies when operating at the CPP-R in Fig. 13 are
significantly higher than the other methods. It can also be
noticed that the P&O-CPG strategies (both at the CPP-R
and CPP-L) have larger errors in Fig. 14 compared to those
in Fig. 13, while the tracking errors of the P- and I-CPG
strategies remain almost at the same level. This increased
tracking error is corresponding to the power overshoot in Fig.
14 as it has been discussed previously. A similar trend is also
observed in Figs. 15 and 16, where it can be seen that the
tracking error of the P&O-CPG method during the clody day
condition is significantly larger than the case during the clear
day condition, while the P- and I-CPG strategies have almost
the same tracking error. Notably, only the tracking error during
stable CPG operation is considered in this case.
D. Stability
Stability is one of the most important aspects for the
CPG control schemes, since the PV system should be able
to continuously deliver power to the grid regardless of the
operating condition. Thus, the presented CPG strategies are
also benchmarked in terms of stability. For the PV systems,
instability may occur during a fast decreasing irradiance con-
dition which can be further divided into two cases related to:
1) short-circuit condition, and 2) open-circuit condition. The
occurrence of the short-circuit instability and its mechanism
have been previously discussed in §II. This type of instability
can occur with the I-CPG strategy where the PV current ipv
is regulated, as it can be observed in Figs. 15(b) and 16(b). In
fact, it can also be seen from the operating trajectories in Fig.
17(b) that the operating point of the PV system almost goes
into the short-circuit condition during a decreasing irradiance
level. Another case of instability is when the operating point
falls into (and stay at) the open-circuit condition. This open-
circuit instability can occur in the case of the P- and P&O-CPG
algorithms when the operating point is chosen at the CPP-R.
The operating point may go into the open-circuit condition
during a decreasing irradiance condition if the PV power is
regulated too far at the right side of the MPP (i.e., at C), since
Fig. 18. Possible operating regions of the CPG strategy, where the instability
issue during the fast decreasing irradiance condition is illustrated.
the open-circuit voltage in the P−Vcurve decreases as the
irradiance level drops (e.g., from 1000 W/m2to 200 W/m2).
The mechanism of the open-circuit instability is illustrated in
Fig. 18 (i.e., C→D). Figs. 16(a) and (c) verify that the P-CPG
or the P&O-CPG at the right side of the MPP can go into
instability during transients. In contrast, it can be seen in Figs.
15 and 16 that the P&O-CPG algorithm can always ensure a
stable operation regardless of the irradiance conditions, only
when the PV system operating point is regulated at the CPP-L.
In this operating region, the sudden drops in the irradiance will
not lead to either the short-circuit or open-circuit instability,
as it can be seen from Fig. 18 (i.e., A→B).
E. Complexity
When comparing all the above CPG strategies, it is found
that the I-CPG algorithm has the simplest control structure,
where only one additional current limiter needs to be added
to the original MPPT controller in Fig. 4(b). Besides, the
calculation of the ilimit is also simple by dividing Plimit by
the measured PV voltage vpv. The control structure of the
P-CPG algorithm is more complicated, basically due to the
MPPT controller in Fig. 4(c). In the case of the P&O-CPG
algorithm, the modification needs to be done at the MPPT
algorithm level as it can be seen from Fig. 10. This makes the
design of a P&O-CPG controller more complicated than the
other two CPG algorithms.
Table II further summarizes a comparison of the results of
the CPG control schemes, in terms of dynamic and steady-state
performances, tracking error, stability, and complexity. The
benchmarking results have validated the effectiveness of the
CPG strategies under various test conditions. It turns out that
the P-CPG strategy can achieve very fast dynamics, especially
during fast changing irradiance condition, compared to the
other strategies. However, this method may induce instability
during the sudden irradiance drops, if the PV system operates
at a low level of power limit (i.e., CPP-R is far away from
MPP). Thus, it is suitable to be implemented in the PV system
with historical fast changing irradiance profiles (e.g., small
scale PV system with cloudy conditions), and a high level of
power limit (i.e., operate at the CPP-R close to the MPP),
in order to minimize the risk of instability. On the other
hand, the P&O-CPG algorithm (when operating at the CPP-

10 IEEE INDUSTRY APPLICATIONS MAGAZINE, VOL. PP, NO. 99, 2016
TABLE II
BENCHMARKING OF THE CON STAN T POWER GENERATION ALGORITHMS.
CPG Strategy Dynamic Responses Steady-state Responses Tracking Error Stability Complexity
Power control (P-CPG) + + + + - -
Current limit (I-CPG) + + + + - - + +
P&O-CPG at CPP-R - - - - - - - -
P&O-CPG at CPP-L - - + + - + + -
Note: the more +, the less tracking error, better stability, and less complexity.
L) is the most suitable approach to realize the CPG control
practically due to its robustness and feasible to be used for
the future grid codes. This method is also suitable when a
wide range of CPG operation (e.g., at different level of power
limit) is required. However, the tracking error of the P&O-CPG
algorithm increases during fast changing irradiance conditions,
which is a trade-off that should be considered.
V. CONCLUSION
In this paper, three Constant Power Generation (CPG)
control solutions for single-phase grid-connected PV systems
have been presented. A benchmarking of the three CPG
control methods has also been conducted in terms of dynamic
and steady-state performances, tracking error, stability, and
complexity. Comparisons have revealed that the CPG strategy
based on a current limit method (I-CPG) has the simplest
control structure. Additionally, the power control based CPG
scheme (P-CPG) has fast dynamics and good steady-state
responses. However, instability may occur in both I-CPG and
P-CPG methods during the operational mode transition, e.g.,
in the case of a fast change in the solar irradiance. It can be
concluded that the CPG based on the P&O algorithm (P&O-
CPG) is the best one in terms of high robustness among the
three CPG strategies once the PV system is operating at the
left side of the maximum power point.
REFERENCES
[1] REN21, “Renewables 2015: Global Status Report (GRS),” 2015.
[Online]. Available: http://www.ren21.net/.
[2] Fraunhofer ISE, “Recent Facts about Photovoltaics in Germany,” May
19, 2015. [Online]. Available: http://www.pv-fakten.de/.
[3] Solar Power Europe, “Global Market Outlook For Solar Power 2015 -
2019,” 2015. [Online]. Available: http://www.solarpowereurope.org/.
[4] Y. Yang, P. Enjeti, F. Blaabjerg, and H. Wang, “Wide-scale adoption
of photovoltaic energy: Grid code modifications are explored in the
distribution grid,” IEEE Ind. Appl. Mag., vol. 21, no. 5, pp. 21–31,
Sep. 2015.
[5] S. Eftekharnejad, V. Vittal, G. Heydt, B. Keel, and J. Loehr, “Impact
of increased penetration of photovoltaic generation on power systems,”
IEEE Trans. Power Syst., vol. 28, no. 2, pp. 893–901, May 2013.
[6] R. Tonkoski, D. Turcotte, and T. El-Fouly, “Impact of high PV pene-
tration on voltage profiles in residential neighborhoods,” IEEE Trans.
Sustain. Energy, vol. 3, no. 3, pp. 518–527, July 2012.
[7] T. Stetz, J. von Appen, F. Niedermeyer, G. Scheibner, R. Sikora, and
M. Braun, “Twilight of the grids: The impact of distributed solar on
germany’s energy transition,” IEEE Power Energy Mag., vol. 13, no. 2,
pp. 50–61, Mar. 2015.
[8] D. Maxwell, “Parts of Northern Ireland’s electricity grid overloaded,”
BBC News NI, 2013. [Online]. Available: http://www.bbc.com/.
[9] E. Reiter, K. Ardani, R. Margolis, and R. Edge, “Industry perspectives
on advanced inverters for us solar photovoltaic systems: Grid benefits,
deployment challenges, and emerging solutions,” NREL (National Re-
newable Energy Laboratory (NREL), Tech. Rep., 2015.
[10] T. Stetz, F. Marten, and M. Braun, “Improved low voltage grid-
integration of photovoltaic systems in Germany,” IEEE Trans. Sustain.
Energy, vol. 4, no. 2, pp. 534–542, Apr. 2013.
[11] A. Ahmed, L. Ran, S. Moon, and J.-H. Park, “A fast PV power tracking
control algorithm with reduced power mode,” IEEE Trans. Energy
Convers., vol. 28, no. 3, pp. 565–575, Sep. 2013.
[12] T. Caldognetto, S. Buso, P. Tenti, and D. Brandao, “Power-based control
of low-voltage microgrids,” IEEE Trans. Emerg. Sel. Topics Power
Electron., vol. 3, no. 4, pp. 1056–1066, Dec. 2015.
[13] German Federal Law: Renewable Energy Sources Act (Gesetz fur den
Vorrang Erneuerbarer Energien) BGBl, Std., Jul. 2014.
[14] Energinet.dk, “Technical regulation 3.2.2 for PV power plants with a
power output above 11 kW,” Tech. Rep., 2015.
[15] Y. Yang, H. Wang, F. Blaabjerg, and T. Kerekes, “A hybrid power control
concept for PV inverters with reduced thermal loading,” IEEE Trans.
Power Electron., vol. 29, no. 12, pp. 6271–6275, Dec. 2014.
[16] H. Beltran, E. Bilbao, E. Belenguer, I. Etxeberria-Otadui, and P. Ro-
driguez, “Evaluation of storage energy requirements for constant pro-
duction in PV power plants,” IEEE Trans. on Ind Electron., vol. 60,
no. 3, pp. 1225–1234, Mar. 2013.
[17] W. A. Omran, M. Kazerani, and M. M. A. Salama, “Investigation
of methods for reduction of power fluctuations generated from large
grid-connected photovoltaic systems,” IEEE Trans. on Energy Convers.,
vol. 26, no. 1, pp. 318–327, Mar. 2011.
[18] A. Sangwongwanich, Y. Yang, F. Blaabjerg, and H. Wang, “Bench-
marking of constant power generation strategies for single-phase grid-
connected photovoltaic systems,” in Proc. of APEC, pp. 370–377, Mar.
2016.
[19] C. Rosa, D. Vinikov, E. Romero-Cadaval, V. Pires, and J. Martins, “Low-
power home PV systems with MPPT and PC control modes,” in Proc.
of CPE, pp. 58–62, Jun. 2013.
[20] W. Cao, Y. Ma, J. Wang, L. Yang, J. Wang, F. Wang, and L. Tolbert,
“Two-stage PV inverter system emulator in converter based power grid
emulation system,” in Proc. of ECCE, pp. 4518-4525, Sep. 2013.
[21] F. Olivier, P. Aristidou, D. Ernst, and T. V. Cutsem, “Active management
of low-voltage networks for mitigating overvoltages due to photovoltaic
units,” IEEE Trans. Smart Grid, vol. 7, no. 2, pp. 926–936, Mar. 2016.
[22] F. Yang, L. Yang, and X. Ma, “An advanced control strategy of PV
system for low-voltage ride-through capability enhancement,” Solar
Energy, vol. 109, pp. 24–35, 2014.
[23] A. Urtasun, P. Sanchis, and L. Marroyo, “Limiting the power generated
by a photovoltaic system,” in Proc. of SSD, pp. 1-6, Mar. 2013.
[24] Y. Chen, C. Tang, and Y. Chen, “PV power system with multi-mode
operation and low-voltage ride-through capability,” IEEE Trans. Ind.
Electron., vol. 62, no. 12, pp. 7524–7533, Dec. 2015.
[25] R.G. Wandhare and V. Agarwal, “Precise active and reactive power
control of the PV-DGS integrated with weak grid to increase PV
penetration,” in Proc. of PVSC, pp. 3150-3155, Jun. 2014.
[26] M. Mirhosseini, J. Pou, and V. Agelidis, “Single- and two-stage inverter-
based grid-connected photovoltaic power plants with ride-through capa-
bility under grid faults,” IEEE Trans. Sustain. Energy, vol. 6, no. 3, pp.
1150–1159, July 2015.
[27] L. D. Watson and J. W. Kimball, “Frequency regulation of a microgrid
using solar power,” in Proc. of APEC, pp. 321–326, Mar. 2011.
[28] A. Sangwongwanich, Y. Yang, and F. Blaabjerg, “High-performance
constant power generation in grid-connected PV systems,” IEEE Trans.
Power Electron., vol. 31, no. 3, pp. 1822–1825, Mar. 2016.
[29] S.B. Kjaer, J.K. Pedersen, and F. Blaabjerg, “A review of single-phase
grid-connected inverters for photovoltaic modules,” IEEE Trans. Ind.
Appl., vol. 41, no. 5, pp. 1292–1306, Sep. 2005.
[30] SMA, “Medium power solutions,” Sunny family 2011/2012, Tech. Rep.

SANGWONGWANICH et al.: BENCHMARKING OF CPG STRATEGIES FOR SINGLE-PHASE GRID-CONNECTED PV SYSTEMS 11
[31] H. Ghoddami and A. Yazdani, “A bipolar two-stage photovoltaic system
based on three-level neutral-point clamped converter,” in Proc. IEEE
Power Energy Soc. Gen. Meet, pp. 1–8, Jul. 2012.
[32] B. Yang, W. Li, Y. Zhao, and X. He, “Design and analysis of a grid-
connected photovoltaic power system,” IEEE Trans. Power Electron.,
vol. 25, no. 4, pp. 992–1000, Apr. 2010.
[33] F. Blaabjerg, R. Teodorescu, M. Liserre, and A.V. Timbus, “Overview
of control and grid synchronization for distributed power generation
systems,” IEEE Trans. Ind. Electron., vol. 53, no. 5, pp. 1398–1409,
Oct. 2006.
[34] N. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, Power electronics
and control techniques for maximum energy harvesting in photovoltaic
systems. CRC press, 2012.
[35] R. W. Erickson and D. Maksimovic, Fundamentals of power electronics,
2nd ed. Norwell, MA: Kluwer, 2001.
[36] N. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli, “Optimization of
perturb and observe maximum power point tracking method,” IEEE
Trans. on Power Electron., vol. 20, no. 4, pp. 963–973, July 2005.
Ariya Sangwongwanich (S’15) was born in
Bangkok, Thailand, in 1991. He received the B.Eng.
degree in electrical engineering from Chulalongkorn
University, Thailand, in 2013, and the M.Sc. in en-
ergy engineering from Aalborg University, Denmark,
in 2015, where he is currently working as a research
assistant. His research interests include control of
grid-connected converter, photovoltaic systems, and
high-power multilevel converters.
Yongheng Yang (S’12-M’15) received the B.Eng.
degree in 2009 from Northwestern Polytechnical
University, China and the Ph.D. degree in 2014 from
Aalborg University, Denmark.
He was a postgraduate with Southeast University,
China, from 2009 to 2011. In 2013, he was a Visiting
Scholar with Texas A&M University, USA. Since
2014, he has been with the Department of Energy
Technology, Aalborg University, where currently he
is an Assistant Professor. His research interests are
focused on grid integration of renewable energy sys-
tems, power converter design, analysis and control, harmonics identification
and mitigation, and reliability in power electronics. Dr. Yang has published
more than 80 technical papers and co-authored a book Periodic Control of
Power Electronic Converters (London, UK: IET, 2017).
Dr. Yang is a Member of the IEEE Power Electronics Society (PELS)
Students and Young Professionals Committee, where he serves as the Global
Strategy Chair and responsible for the IEEE PELS Students and Young
Professionals Activities. He served as a Guest Associate Editor of IEE E
JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER
ELECT RONICS, and has also been invited as a Guest Editor of Applied
Sciences. He is an active reviewer for relevant top-tier journals.
Frede Blaabjerg (S’86-M’88-SM’97-F’03) was
with ABB-Scandia, Randers, Denmark, from 1987
to 1988. From 1988 to 1992, he was a Ph.D.
Student with Aalborg University, Aalborg, Denmark.
He became an Assistant Professor in 1992, As-
sociate Professor in 1996, and Full Professor of
power electronics and drives in 1998. His current
research interests include power electronics and its
applications such as in wind turbines, PV systems,
reliability, harmonics and adjustable speed drives.
He has received 17 IEEE Prize Paper Awards, the
IEEE PELS Distinguished Service Award in 2009, the EPE-PEMC Council
Award in 2010, the IEEE William E. Newell Power Electronics Award 2014
and the Villum Kann Rasmussen Research Award 2014. He was an Editor-in-
Chief of the IEEE TRANSACTIONS ON POWER ELECTRONICS from
2006 to 2012. He is nominated in 2014 and 2015 by Thomson Reuters to be
between the most 250 cited researchers in Engineering in the world.
Huai Wang (S’07-M’12) is currently an Associate
Professor and a research trust leader with the Center
of Reliable Power Electronics (CORPE), Aalborg
University, Denmark. His research addresses the fun-
damental challenges in modelling and validation of
failure mechanisms of power electronic components,
and application issues in system-level predictability,
circuit architecture, and robustness design. Dr. Wang
received his PhD degree from the City University of
Hong Kong, Hong Kong, in 2012. He has served as
Associate Editor of IEEE TRAN SAC TIONS O N
POW ER ELECT RON ICS since 2015. He received the IEEE PELS Richard
M. Bass Outstanding Young Power Electronics Engineer Award, in 2016.