ArticlePDF Available

Experimental validation of a battery dynamic model

Authors:

Abstract and Figures

This paper presents an improved and easy-to-use battery dynamic model. The charge and the discharge dynamics of the battery model are validated experimentally with four batteries types. An interesting feature of this model is the simplicity to extract the dynamic model parameters from batteries datasheets. Only three points on the manufacturer's discharge curve in steady state are required to obtain the parameters. Finally, the battery model is included in the SimPowerSystems simulation software and used in a detailed simulation of an electric vehicle based on a hybrid fuel cell-battery power source. The results show that the model can accurately represent the dynamic behaviour of the battery.
Content may be subject to copyright.
EVS24
Stavanger, Norway, May 13 - 16, 2009
Experimental Validation of a Battery Dynamic Model for EV
Applications
Olivier Tremblay1, Louis-A. Dessaint
Electrical Engineering Department, ´
Ecole de Technologie Sup ´
erieure
1Email: olivier.tremblay.1@ens.etsmtl.ca
Abstract
This paper presents an improved and easy-to-use battery dynamic model. The charge and the discharge dynamics
of the battery model are validated experimentally with four batteries types. An interesting feature of this model
is the simplicity to extract the dynamic model parameters from batteries datasheets. Only three points on the
manufacturer’s discharge curve in steady state are required to obtain the parameters. Finally, the battery model
is included in the SimPowerSystems simulation software and used in a detailed simulation of an electric vehicle
based on a hybrid fuel cell-battery power source. The results show that the model can accurately represent the
dynamic behaviour of the battery.
Keywords: Simulation; Battery Model; SimPowerSystems; Fuel Cell Vehicle
1 Introduction
The central element of electric vehicles (EV) and of
more electric systems in general is the battery. This
element stores a great amount of energy to be re-
lease when necessary. The battery enables regenera-
tive braking in an EV and allows to supplement a slow
dynamic energy source, such as the fuel cell. Bene-
fits are drawn from these features in EVs based on fuel
cells such as the Honda FCX Clarity. The battery’s
management system (BMS) must ensure an efficient
management battery’s state of charge (SOC). To ac-
complish this, the designer of the BMS must have a
detailed simulation of the EV’s traction system includ-
ing a detailed model of the battery.
There are basically three types of battery models re-
ported in the literature, specifically: experimental,
electrochemical and electric circuit-based. Experi-
mental and electrochemical models are not well suited
to represent cell dynamics for the purpose of state-of-
charge estimations of battery packs. However, elec-
tric circuit-based models can be useful to represent
electrical characteristics of batteries. The most sim-
ple electric model consists of an ideal voltage source
in series with an internal resistance [2]. This model,
however, does not take into account the battery SOC.
There is another model based on an open circuit volt-
age in series with a resistance and parallel RC circuits
with the so-called Warburg impedance [3]. The iden-
tification of all the parameters of this model is based
on a rather complicated technique called impedance
spectroscopy [4]. Shepherd developed an equation to
describe the electrochemical behaviour of a battery di-
rectly in terms of terminal voltage, open circuit volt-
age, internal resistance, discharge current and state-of-
charge [1], and this model is applied for discharge as
well as for charge. A modified version of the Shepherd
model has been used in [5]. This modification con-
sists on using a polarisation voltage instead of a po-
larisation resistance in order to remove the algebraic
loop problem due to the simulation of electrical sys-
tems in Simulink. This model uses only the battery
SOC as a state variable to represent the voltage be-
haviour. This is valid in steady state (constant current)
but this model produces false results when the current
varies. The main contribution of this paper is to show
how the SimPowerSystems battery model was recently
improved in order to extend its validity for variable
charging or discharging current.
The rest of this paper is organized as follows: In sec-
tion 2, the mathematical model which varies accord-
ing to the battery type will be described. Section 3
will present the model’s parameters and the procedure
to deduct them from the battery manufacturer’s dis-
charge curve. Section 4 will present an interesting
contribution of this paper as it shows the results of an
experimental validation for four types of batteries. Fi-
nally, section 5 will present an application of the bat-
tery model in a full simulation of the EV traction sys-
tem based on a fuel cell.
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 1
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0289
Figure 1: Typical discharge curve
2 The proposed battery model
2.1 The discharge model
The Fig. 1 shows a typical discharge characteristic, for
a Nickel-Metal-Hydrid cell.
The proposed discharge model is similar to the Shep-
herd model but can represent accurately the voltage
dynamics when the current varies and takes into ac-
count the open circuit voltage (OCV) as a function
of SOC. A term concerning the polarisation voltage
is added to better represent the OCV behaviour and
the term concerning the polarisation resistance [1] is
slightly modified. The battery voltage obtained is
given by:
Vbatt =E0KQ
Qit ·it
 
P ol. V oltage
R·i+
Aexp(B·it)KQ
Qit·
 
P ol. Resistance
i
(1)
where
Vbatt = battery voltage (V)
E0= battery constant voltage (V)
K= polarisation constant (V/(Ah)) or polarisation
resistance (Ω)
Q= battery capacity (Ah)
it =idt = actual battery charge (Ah)
A= exponential zone amplitude (V)
B= exponential zone time constant inverse (Ah)1
R= internal resistance (Ω)
i= battery current (A)
i= filtered current (A)
Figure 2: Hysteresis phenomenon
The particularity of this model is the use of a fil-
tered current (i) flowing through the polarisation re-
sistance. In fact, experimental results show a voltage
slow dynamic behaviour for a current step response.
This filtered current solve also the algebraic loop prob-
lem due to the simulation of electrical systems in
Simulink. Finally, the OCV varies non-linearly with
the SOC. This phenomenon is modelled by the polari-
sation voltage term.
The exponential zone of equation (1) is valid for the
Li-Ion battery. For the other batteries (Lead-Acid,
NiMH and NiCD), there is a hysteresis phenomenon
between the charge and the discharge, no matter the
SOC of the battery [6], [7]. This behaviour occurs only
in the exponential area, as shown in Fig. 2.
This phenomenon can be represented by a non-linear
dynamic system:
˙
Exp(t)=B·|i(t)(Exp(t)+A·u(t)) (2)
where
Exp(t)= exponential zone voltage (V)
i(t)= battery current (A)
u(t)= charge or discharge mode
The exponential voltage depends on its initial value
Exp(t0)and the charge (u(t) = 1) or discharge (u(t) =
0) mode. Fig. 3 shows the complete discharge model
system.
2.2 The charge model
The charge behaviour, particularly the end of the
charge (EOC) characteristic, is different and depends
on the battery type.
2.2.1 Lead-Acid and Li-Ion batteries
The Lead-Acid and Li-Ion batteries have the same
EOC characteristics, because the voltage increases
rapidly when the battery reaches the full charge. This
phenomenon is modelled by the polarisation resistance
term. In the charge mode, the polarisation resistance
increases until the battery is almost fully charged (it
= 0). Above this point, the polarisation resistance in-
creases abruptly.
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 2
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0290
Figure 3: Discharge battery model
Instead of the polarisation resistance of the discharge
model (equation (1)), the polarisation resistance is now
given by:
P ol. Resistance =KQ
it (3)
Theoretically, when it = 0 (fully charged), the polarisa-
tion resistance is infinite. This is not exactly the case in
practice. Indeed, experimental results have shown that
the contribution of the polarisation resistance is shifted
by about 10% of the capacity of the battery. Equation
(3) can be re-written to:
P ol. Resistance =KQ
it 0.1·Q(4)
Similar to the discharge model, the exponential volt-
age for the Li-Ion battery is the A·exp(B·it)term
and for the Lead-Acid, the voltage is given by equation
(2).
2.2.2 NiMH and NiCd batteries
These two types of batteries have a particular be-
haviour at EOC. Indeed, after the battery has reached
the full charge voltage, the voltage decreases slowly,
depending on the current amplitude. This phe-
nomenon is very important to model because a battery
charger monitors the ΔVvalue to stop the charge.
This behaviour is represented by modifying the charge
polarisation resistance. When the battery is fully
charged (it =0), the voltage starts to drop. The
charger continues to overcharge the battery (it < 0)
and the voltage decreases. This phenomenon can be
represented by decreasing the polarisation resistance
when the battery is overcharged by using the absolute
value of the charge (it):
P ol. Resistance =KQ
|it|−0.1·Q(5)
Similar to the discharge model, the exponential volt-
age for these batteries is given by equation (2).
2.2.3 The charge model overview
The complete charge model is similar to the one in Fig.
3. Only the Ebatt calculation differs, depending on the
battery type:
Lead-Acid : Ebatt =E0KQ
it0.1·Q·i
KQ
Qit ·it +Exp(t)
Li-Ion : Ebatt =E0KQ
it0.1·Q·iKQ
Qit ·
it +Aexp(B·it)
NiMH and NiCd : Ebatt =E0KQ
|it|−0.1·Q·
iKQ
Qit ·it +Exp(t)
2.3 Model summary
2.3.1 Lead-Acid
Discharge : Vbatt =E0R·iKQ
Qit ·(it +
i)+Exp(t)
Charge : Vbatt =E0R·iKQ
it0.1·Q·i
KQ
Qit ·it +Exp(t)
2.3.2 Li-Ion
Discharge : Vbatt =E0R·iKQ
Qit ·(it +
i)+Aexp(B·it)
Charge : Vbatt =E0R·iKQ
it0.1·Q·i
KQ
Qit ·it +Aexp(B·it)
2.3.3 NiMH and NiCd
Discharge : Vbatt =E0R·iKQ
Qit ·(it +
i)+Exp(t)
Charge : Vbatt =E0R·iKQ
|it|−0.1·Q·i
KQ
Qit ·it +Exp(t)
The proposed model is based on specific assumptions
and has limitations:
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 3
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0291
Type Lead- NiCd Li-Ion NiMH
Acid
Parameters 12V 7.2Ah 1.2V 2.3Ah 3.3V 2.3Ah 1.2V 6.5Ah
E0(V)12.4659 1.2705 3.366 1.2816
R(Ω) 0.04 0.003 0.01 0.002
K(Ωor V/(Ah)) 0.047 0.0037 0.0076 0.0014
A (V) 0.83 0.127 0.26422 0.111
B(Ah)1125 4.98 26.5487 2.3077
Table 1: Battery parameters
2.3.4 Model assumptions
The internal resistance is supposed constant dur-
ing the charge and discharge cycles and does not
vary with the amplitude of the current.
The model’s parameters are deduced from the
discharge characteristics and assumed to be the
same for charging.
The capacity of the battery does not change with
the amplitude of the current (No Peukert effect).
The temperature does not affect the model’s be-
haviour.
The Self-Discharge of the battery is not repre-
sented.
The battery has no memory effect.
2.3.5 Model limitations
The minimum No-Load battery voltage is 0 V and
the maximum battery voltage 2*E0.
The minimum capacity of the battery is 0 Ah and
the maximum capacity Q. Therefore, the maxi-
mum SOC cannot be greater than 100% if the bat-
tery is overcharged.
3 Extracting the parameters
An important feature of the proposed model is the sim-
plicity with which the dynamic model parameters are
extracted. In fact, it is not necessary to take experi-
mental measures on the battery in order to extract the
parameters. Only three points on the manufacturer’s
discharge curve, in steady state, are required to ob-
tain the parameters. Battery manufacturers provide
datasheet which includes a ”Typical Discharge Char-
acteristics” curve (Fig. 1) where it is possible to ex-
tract the fully charged voltage (Vfull), the end of the
exponential zone (Qexp, Vexp), the end of the nom-
inal zone (Qnom, Vnom) (when the voltage starts to
drops abruptly) and the maximum capacity (Q). Also,
the internal resistance (R) is generally given.
With these three points, it is possible to solve, using
equation 1, the following set of equations (equations 6,
7 and 8). Note that the manufacturer curve is obtained
at constant current (generally equal to 0.2C).
For the fully charged voltage, the extracted charge is 0
(it =0) and the filtered current (i) is 0 because the
current step has just started:
Vfull =E0R·i+A(6)
For the end of the exponential zone, the factor B can
be approximated to 3/Qexp since the energy of the ex-
ponential term is almost 0 (5 %) after 3 time constants.
The filtered current (i) is equal to ”i” because the cur-
rent is in steady state:
Vexp =E0KQ
QQexp ·(Qexp +i)
R·i+Aexp(3
Qexp ·Qexp)
(7)
The nominal zone voltage is given by:
Vnom =E0KQ
QQnom ·(Qnom +i)
R·i+Aexp(3
Qexp ·Qnom)
(8)
Finally, the time constant of the filtered current (i)
is not given by the manufacturer datasheet. Only ex-
perimental test can provide this information. However,
experimental data have shown a time constant of about
30 secs for the four batteries types.
This general approach can be applied to other battery
types to obtain the model parameters. Obviously, these
parameters are approximate and the level of accuracy
of the model depends on the precision of the points
extracted from the discharge curve.
The above approach has been used to extract param-
eters for Lead-Acid, Nickel-Cadmium, Nickel-Metal-
Hydride and Lithium-Ion batteries. The model param-
eters found for common battery cells are presented in
Table 1.
4 The model validation
The proposed model is first validated at steady state
in order to reproduce the manufacturers ”Typical Dis-
charge Characteristics” curve. Since most of them do
not provide the battery dynamic performance, exper-
imental tests are performed on the battery to demon-
strate the validity of the model.
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 4
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0292
Figure 4: Simulation and datasheet results
4.1 Steady state validation
A model of a 6.5Ah, 1.2V, NiMH battery (HHR650D
from panasonic)1is obtained using its datasheet:
Internal resistance [R] = 0.002 Ω
Discharge current [i] = 1.3 A
Maximum capacity [Q]=7Ah(5.38h x 1.3A)
Fully charged voltage [Vfull] = 1.39 V
End of the exponential zone [Qexp,Vexp] = [1.3
Ah, 1.28 V]
Nominal zone [Qnom,Vnom] = [6.25 Ah, 1.18 V]
Fig. 4 shows the simulation results superimposed on
the datasheet curve. The first graph shows the volt-
age vs time for a 0.2C current and the second graph
shows the voltage for 1C, 2C and 5C currents. The
blue solid line represents the simulation results. It is
observed that the simulated curves match very well
the real curves during almost 90% of the discharge,
no matter the discharge current amplitude.
1The datasheet can be found at http :
//www.panasonic.com/industrial/battery
/oem/chem/nicmet
Figure 9: Dynamic charge-discharge of a 2.3Ah, 3.3V Li-Ion
battery
4.2 Validation of dynamic behaviour
An interesting feature of this paper is the validation of
the battery’s dynamic behaviour with respect to cur-
rent variation and the battery’s SOC. The validation
has been done using chart recorder and a current con-
trolled load. A contact resistance has been added to the
simulation model to represent the voltage drop caused
by the battery terminals.
Figs. 5, 6, 7 and 8 show the simulation results super-
imposed to the experimental results for the four batter-
ies types. The parameters of these batteries are those
found in Table 1. The first graph shows the experimen-
tal battery voltage (black solid line) and the simulated
battery voltage (blue dotted line). The second graph
represents the battery current. The third graph shows
the estimated battery SOC. Finally, the absolute error
(in percent) between the measured and simulated volt-
age is shown in the fourth graph.
It can be noted that the error between the simulated
voltage and the real voltage is within ±5% for SOC
between 100 % and 20 % during the charge and the
discharge mode. When the SOC decreases below 20
%, the precision of the simulation model is around
±10%. This is quite acceptable because it is not rec-
ommended to fully discharge a battery. However, the
Lead-Acid battery model has a validity domain be-
tween 100 % and 30 % (SOC) because the Peukert
effect is not taken into account.
A critical test for the battery model is to switch be-
tween the charge and the discharge modes in order
to verify the model’s continuity. Fig. 9 shows four
discharge-charge cycles for the Li-Ion battery. Results
show that the model has a precision within ±3%.
Another important point is the validation of the hys-
teresis effect for the NiMH battery. Fig. 10 shows one
charge-discharge cycle for a 2 Ah, 1.2V NiMH battery.
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 5
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0293
(a) Discharge (b) Charge
Figure 5: Dynamic discharge and charge of a 6.5Ah, 1.2V NiMH battery
(a) Discharge (b) Charge
Figure 6: Dynamic discharge and charge of a 2.3Ah, 3.3V Li-Ion battery
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 6
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0294
(a) Discharge (b) Charge
Figure 7: Dynamic discharge and charge of a 2.3Ah, 1.2V NiCd battery
(a) Discharge (b) Charge
Figure 8: Dynamic discharge and charge of a 7.2Ah, 12V Lead-Acid battery
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 7
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0295
Figure 10: Dynamic charge-discharge of a 2.0Ah, 1.2V
NiMH battery
It can be noted that at the beginning of the simulation
(when the charge current of 4A is applied), the mea-
sured voltage is 1.543V and the simulated voltage is
1.471V. The corresponding error is 4.7 %. During the
charge period, the error is very low. When the mode
changes from the charge to discharge (at t = 600 secs),
a maximum error of 5 % is noted and this error in-
creases to9%inthedischarge zone. An important
point to note occurs at t = 1200 secs, where the current
becomes negative (charge current). At this time the
measured voltage is 1.476V and the simulated voltage
is 1.396V, for a corresponding error of 5 %. The differ-
ence between the voltage at t = 0 sec and t = 1200 secs
represents the hysteresis phenomenon (0.067V for the
experimental results and 0.075V for the simulated re-
sults). From these results, it can be stated that the
hysteresis effect is well represented by the simulation
model.
This experimental validation has shown that even if
the model parameters are extracted from a steady state
curve, it is possible to simulate accurately (error within
±5%) the dynamic behaviour of the battery for the
charge and the discharge processes. The validity do-
main of the model is included between 100 % and 20
% of SOC for the NiMH, NiCd and Li-Ion batteries
and between 100 % and 30 % for the Lead-Acid bat-
tery because the Peukert effect is not modelled. Also,
the model behaviour when the current changes sud-
denly from charge to discharge is coherent with the ex-
perimental results. Finally, the model represents well
the hysteresis phenomenon for the NiMH battery (the
model represents also the effect for the NiCd battery
and the small effect appearing with the Lead-Acid bat-
tery).
Figure 11: Fuel Cell Vehicle simulation model
5 Application to a Fuel Cell Elec-
tric Vehicle
The new model is now integrated in the Matlab-
Simulink SimPowerSystems R2009a library. A user-
friendly interface allows the user to enter standardized
parameters, and then the model’s parameters are cal-
culated automatically according to a similar method to
that presented in section 3. The user can then visual-
ize the discharge curve obtained with the parameters
and compare it with that of the manufacturer. There
are four sets of preset parameters making it possible to
represent the behaviour of the batteries determined in
Table 1. Of course if desired, it is possible to refine the
parameters for a particular battery behaviour.
The model is used in a detailed multi-domain simula-
tion of a EV traction system based on a fuel cell (Fig.
11). The multi-domain simulation allows for the de-
sign and fine tuning of the vehicle’s energy manage-
ment system in order to accurately share the electri-
cal power between the two sources of energy and the
permanent magnet synchronous motor. The Matlab-
Simulink blockset series allows to model, in an unique
simulation environment, the electrical and mechanical
systems and the various control systems. The fuel cell
vehicle (FCV) is based on a recent topology, such as
the Honda FCX Clarity 2008. The simulation model
is composed of three main modules:
The FCV Electrical Subsystem (Fig. 12) con-
tains a 100 kW, 288 Vdc Interior Permanent Mag-
net Synchronous Motor (IPMSM) with the asso-
ciated drive, achieving a maximum motor speed
of 12500 rpm. The drive is torque regulated and
the reference (T
motor)comes from the energy
management system. The electrical power comes
mainly from a 100 kW Proton Exchange Mem-
brane Fuel Cell (PEMFC). The reference current
(I
FC)is used by the fuel cell system to feed the
stack in Hydrogen and Oxygen. Also, this ref-
erence current is used by the DC/DC converter
(used to interface the fuel cell voltage to the DC
bus voltage) to regulate the current fed to the DC
bus. During power transients, the fuel cell stack
cannot meet the required power due to its large
time constant. Therefore, a Li-Ion battery (288
Vdc, 13.9 Ah, 75 kW Max) is used to assist the
fuel cell and to restore the vehicle braking’s en-
ergy. Also, the battery is used to re-accelerate the
vehicle when the energy is available.
The Energy Management Subsystem (Fig. 13)
represents the key part in this simulation. In fact,
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 8
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0296
Figure 12: Fuel Cell Vehicle electrical system
Figure 13: Energy Management System
it determines the required torque of the electrical
motor, and the required current of the fuel cell,
according to the maximum battery power and the
battery SOC. The required vehicle power depends
on the pedal position, which varies between -100
% (full brake) and 100 % (full acceleration). This
subsystem contains also a battery management
system which ensure that the battery SOC is al-
ways between 40 % and 80 %.
The FCV Vehicle Dynamics models, with the
SimDriveline blockset, all the mechanical com-
ponents: reduction gear, differential, tires dynam-
ics and vehicle dynamics.
More particularly, the section concerning the EMS,
shown in Fig. 13, is studied. The EMS ensures that
the required motor power never exceed the available
power. Depending on the pedal position and the vehi-
cle speed, the required drive power (P
drive )is calcu-
lated. The required fuel cell power (P
FC)is obtained
by subtracting the battery recharge power (P
recharge )
to P
drive . A lookup table, based on the polarisation
curve, is used to calculate the required fuel cell cur-
rent. As the fuel cell response time is large, the re-
quired fuel cell power cannot be obtained instanta-
neously. The required motor power (P
motor)must
be limited by the total available power (measured fuel
cell power and battery required power) in order to pre-
vent bus voltage collapse. The battery required power
(P
batt)is obtained by subtracting the required fuel cell
power P
FC, to the measured fuel cell power (PFC).If
Figure 14: Energy Management System results
required, (depending on the battery SOC and the max-
imum power) the battery required power is limited.
The FCV is simulated during 16 s and covers four dif-
ferent operation modes: acceleration, cruising, fast ac-
celeration and deceleration. Fig. 14 shows the simula-
tion results:
Att=0s,theFCVisstopped and the driver
pushes the accelerator pedal to 70 %. Since the
fuel cell cannot start as quickly as required, the
battery provides the motor power until the fuel
cell starts.
At t = 0.7 s, the fuel cell begins to provide power
but is not able to reach the reference power due to
its large time constant. Therefore the battery con-
tinues to provide electrical power to the motor.
At t = 4 s, the accelerator pedal is released to 25
%. The fuel cell cannot decrease its power instan-
taneously; therefore the battery absorbs excess of
the fuel cell power in order to maintain the re-
quired power.
Att=6s,thefuel cell power is equal to the ref-
erence power. The battery is no more needed.
At t = 8 s, the accelerator pedal is pushed to 85
%. The battery helps the fuel cell by providing an
extra power of 50 kW.
At t = 9 s, the fuel cell power reaches nearly the
reference power of 100 kW. The battery power is
progressively reduced to 5 kW.
At t = 12 s, the accelerator pedal is set to -25 %,
corresponding to a regenerative power of 50 kW.
The motor acts as a generator driven by the ve-
hicles wheels. The kinetic energy of the FCV is
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 9
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0297
transformed into electrical energy which is stored
in the battery. For this pedal position, the battery
absorbs the regenerative power and the residual
fuel cell power.
At t = 15 s, the fuel cell power is about 2 kW (the
minimum power).
Dynamic simulation is one of the first steps in the de-
sign of systems such as the FCV. The improvement
of the EMS as well as the dimensioning of the com-
ponents can be obtained when the simulation is done
with great exactitude. For example, so as not to exceed
the maximum power of the battery, it is possible to di-
rectly control the electric motor in order to limit its
current. These current peaks can be determined with
precision through dynamic simulation.
Simulation using the battery model thus makes it pos-
sible to analyse very complex phenomena. In this
case, the battery’s model parameters are not exactly
the same as those of the Honda FCX Clarity battery
but it is nevertheless possible to study, with good pre-
cision, phenomena caused by the battery. Moreover,
this model helps to develop the EMS, which controls
all the converters as well as the fuel cell, in order not
to exceed the maximum available power. Finally, it is
possible to control the charge and the discharge of the
battery with precision. This model is thus the central
point of the EV’s components since all the other sys-
tems depend on the battery behaviour.
Although little technical information is available on
the Honda FCX Clarity2, it has been possible to vali-
date the simulation results using known variables such
as the 0-100 km/h acceleration, the maximum speed
and the hydrogen consumption:
Fuel consumption of 63.82 miles/kg-H2
The car makes 0-100 km/h in 9s
The maximum car speed when the accelerator is
100 % is 165 km/h)
6 Conclusion
In conclusion, this paper demonstrates that the new
SimPowerSystems battery model allows for an ad-
equate representation of a battery’s real behaviour
based on only three points on the battery manufac-
turer’s discharge curve. It has been demonstrated that
even if the points are extracted from a constant-current
discharge curve, the dynamic behaviour obtained in
simulation is close to the experimental behaviour. Fi-
nally, the integration of the new battery model in a
multi-domain simulation of an EV based on a fuel
cell enables to design and adequately adjust the en-
ergy management system as well as the battery’s man-
agement system. The obtained results are coherent
with reality and the vehicle’s total energy consump-
tion concords with the public information available on
the Honda FCX Clarity.
2From the Honda FCX Clarity Press Kit at
http://www.hondanews.com, the values are 68 miles/ kg-H2,
0-100 km/h in 9.2s and 160 km/h respectively.
References
[1] Shepherd, C. M., Design of Primary and Secondary
Cells - Part 2. An equation describing battery discharge,
Journal of Electrochemical Society, Volume 112, July
1965, pp 657-664.
[2] Durr, Matthias; Cruden, Andrew; Gair, Sinclair; Mc-
Donald, J.R, Dynamic model of a lead acid battery for
use in a domestic fuel cell system, Journal of Power
Sources, Volume 161, no 2, Oct 27, 2006, pp. 1400-
1411.
[3] Kuhn, E.; Forgez, C.; Lagonotte, P.; Friedrich, G., Mod-
elling Ni-MH battery using Cauer and Foster structures,
Journal of Power Sources, v 158, no 2 SPEC. ISS., Aug
25, 2006, pp. 1490-1497.
[4] Mauracher, P.; Karden, E., Dynamic modelling of
lead/acid batteries using impedance spectroscopy for
parameter identification, Journal of Power Sources, v
67, no 1-2, July/August, 1997, pp. 69-84.
[5] Tremblay, O.; Dessaint, L.-A.; Dekkiche, A.-I., A
Generic Battery Model for the Dynamic Simulation of
Hybrid Electric Vehicles, Vehicle Power and Propulsion
Conference, 2007. VPPC 2007. IEEE , pp. 284-289, 9-
12 Sept. 2007.
[6] Rynkiewicz, R., Discharge and charge modeling of lead
acid batteries, Applied Power Electronics Conference
and Exposition, 1999. APEC ’99. Fourteenth Annual ,
vol.2, no., pp.707-710 vol.2, 14-18 Mar 1999
[7] Feng Xuyun,; Sun Zechang,, A battery model including
hysteresis for State-of-Charge estimation in Ni-MH bat-
tery, Vehicle Power and Propulsion Conference, 2008.
VPPC ’08. IEEE , vol., no., pp.1-5, 3-5 Sept. 2008
Authors
Olivier Tremblay received the B.Ing. and M.Ing. degrees
from the ´
Ecole de Technologie Sup´
erieure, Montr´
eal, Qc,
Canada in 2004 and 2006, respectively, all in electrical en-
gineering. He is currently a research associate at the ´
Ecole de
Technologie Sup´
erieure, Montr´
eal. He is in charge of devel-
oping new simulation models and solvers in the SimPower-
Systems (SPS). He integrated in SPS the models of the elec-
tric battery and the fuel cell and he is author of two important
demos of SPS: the multi-domain simulations of the power
trains of an HEV and of an electric car based on a fuel cell.
Louis-A. Dessaint received the B.Ing., M.Sc.A., and Ph.D.
degrees from the ´
Ecole Polytechnique de Montr´
eal, QC,
Canada, in 1978, 1980, and 1985, respectively, all in electri-
cal engineering. He is currently a Professor of electrical en-
gineering at the ´
Ecole de Technologie Sup´
erieure, Montr´
eal.
From 1992 to 2001, he was the Director of the Groupe de
recherche en ´
electronique de puissance et commande indus-
trielle (GREPCI), a research group on power electronics and
digital control. Since 2002, he has been the Trans ´
Energie
(Hydro-Qu´
ebec) Chair on Power Systems Simulation and
Control. He is one of the authors of the SimPowerSystems
simulation software of MathWorks.
EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 10
World Electric Vehicle Journal Vol. 3 - ISSN 2032-6653 - © 2009 AVERE
Page 0298
... Convective between the walls' external surface and the external air [34] 0 Convective between the walls' external surface and the air inside the driver's cabin [34] 0 Incident solar radiation [34] Radiative with the celestial vault (the vehicle is considered a small convex object placed inside a cavity) [34] 0 Convective between the internal air and the surface of the inner wall bordering the outside Convective between the internal air and the walls' inner surface bordering the driver's cabin Internal due to the auxiliaries Door opening (≠0 only in the goods loading/unloading phases, known duration) [35] , , ...
... Convective between the walls' external surface and the external air [34] 0 Convective between the walls' external surface and the air inside the driver's cabin [34] 0 Incident solar radiation [34] Radiative with the celestial vault (the vehicle is considered a small convex object placed inside a cavity) [34] 0 Convective between the internal air and the surface of the inner wall bordering the outside Convective between the internal air and the walls' inner surface bordering the driver's cabin Internal due to the auxiliaries Door opening (≠0 only in the goods loading/unloading phases, known duration) [35] , , ...
... Convective between the walls' external surface and the external air [34] 0 Convective between the walls' external surface and the air inside the driver's cabin [34] 0 Incident solar radiation [34] Radiative with the celestial vault (the vehicle is considered a small convex object placed inside a cavity) [34] 0 Convective between the internal air and the surface of the inner wall bordering the outside Convective between the internal air and the walls' inner surface bordering the driver's cabin Internal due to the auxiliaries Door opening (≠0 only in the goods loading/unloading phases, known duration) [35] , , ...
Article
Full-text available
The path towards decarbonization requires a progressive adaptation of all refrigeration systems, but only stationary ones have been intensely studied to improve their environmental performance. However, refrigerated transport is vital in the cold chain and must be considered in the green transition. In this paper, we propose a model for a hybrid refrigerated van that includes photovoltaic panels and electric batteries to decrease total greenhouse gas emissions from the engine. Thermal, electrical, and battery sub-models are considered and integrated into the comprehensive hybrid solar-powered refrigerated van model. Different technologies are compared, including lithium and lead-acid batteries and three different types of photovoltaic panels. The model was validated regarding van fuel consumption, showing a 4% deviation. Single and multiple delivery scenarios are considered to assess the energy, economic, and environmental benefits. Monthly CO2,e emissions could be reduced by 20% compared to a standard refrigerated van. Despite the environmental benefits provided by this sustainable solution, the payback period is still too long (above 20 years) because of the necessary investment to adapt the vehicle and considering fuel and electricity prices currently.
... The term E bat depends on the charging and discharging mode of operation of the battery. In this study, lithium-ion battery has been considered and its internal voltage can be written as [33,34] E Charge bat ...
Article
Full-text available
The standalone microgrids with renewable energy resources (RERs) such as a photovoltaic (PV) system and fast changing loads face major challenges in terms of reliability and power management due to a lack of inherent inertial support from RERs and their intermittent nature. Thus, energy storage technologies such as battery energy storage (BES) are typically used to mitigate the power fluctuations and maintain a power balance in the system. This paper presents a model predictive control (MPC) based power management strategy (PMS) for such standalone PV/battery systems. The proposed method is equipped with an autoregressive integrated moving average (ARIMA) prediction method to forecast the load and environmental parameters. The proposed controller has the capabilities of (1) effective power management, (2) minimization of transients during disturbances, and (3) automatic switching of the operation of the PV between the maximum power point tracking (MPPT) mode and power-curtailed mode that prevents the overcharging of the battery and at the same time maximize the PV utilization. The effectiveness of the proposed method has been verified through a comprehensive simulation-based analysis.
Article
Due to the combination of aerodynamic and buoyant lift, a multi-lobes hybrid air vehicle (HAV) is very suitable for cargo transportation, communications and surveillance in the stratosphere. Solar cells can be used as an ideal energy source for the HAV because of its flat upper surface and sufficient solar radiation in the stratosphere. A hybrid energy system that combines solar cells, fuel cells and lithium batteries is considered to power the HAV in this paper. To better manage the various energy subsystems and achieve optimal flight performance, a power-following management strategy, which is a typical rule-based energy management strategy(EMS), is constructed for the hybrid energy system of the HAV. The EMS takes into account the high specific power of lithium batteries and the high specific energy of fuel cells to improve the endurance of HAV while ensuring the maneuverability during takeoff and landing. The simulation results show that output of each energy subsystem not only meets the power demand of the HAV, but also conforms to output characteristics of each subsystem. Furthermore, the simulation analysis is also carried out to investigate the impact of one of the energy failures on energy management and corresponding strategies are given. The results show the advantages of EMS proposed in this paper, which can provide a reference for the design of HAV.
Article
Background: A cost-effective solution for the design of distributed energy storage systems implies the development of battery performance models yielding a suitable representation of its dynamic behaviour under realistic operation conditions. Methods: In this work, a lithium-ion battery (LIB) is tested to be further modelled and integrated into an existing energy management control system. This specific LIB (5.0 kW /9.8 kWh) is integrated with a commercial inverter and solar photovoltaic (PV) system (3.3 kWp) as part of a microgrid that is also encompassing other energy storage technologies at the University of Évora, Pole of INIESC – National Research Infrastructure for Solar Energy Concentration. A testing protocol fully characterizes the battery and the inverter efficiency to describe their performance better. Then, a battery model is built upon both the existent LIB description and experimental fitting regression. The model allows obtaining the voltage curve, the internal resistance (i.e., to describe instantaneous voltage drop/rise and transients), and the state of charge (SOC) and/or energy capacity based on the current input. The developed model is validated through the comparison with the experimental results. Results: The model approach presented a higher voltage RMSE (root mean square error) of 5.51 V and an MRE (maximum relative error) of 5.68 % in the discharge state. Regarding SOC, the MRE obtained was approximately 6.82 %. In the charge state, the highest RMSE voltage was 5.27 V, with an MRE of 6.74 %. Concerning SOC, the MRE obtained was approximately 6.53 %. Conclusions: The developed model is validated through the comparison with experimental results. Based on computational effort, simplicity of use and the associated model error, the approach is validated to the regular conditions of the commercial battery pack to be incorporated in the next research step, following a bottom-up modelling approach for an increasingly more complex smart grid.
Article
Full-text available
This paper deals with dynamic models of Ni-mH battery and focuses on the development of the equivalent electric models. We propose two equivalent electric models, using Cauer and Foster structures, able to relate both dynamic and energetic behavior of the battery. These structures are well adapted to real time applications (e.g. Battery Management Systems) or system simulations. A special attention will be brought to the influence of the complexity of the equivalent electric scheme on the precision of the model. Experimental validations allow to discuss about performances of proposed models.
Conference Paper
Full-text available
This paper presents an easy-to-use battery model applied to dynamic simulation software. The simulation model uses only the battery State-Of-Charge (SOC) as a state variable in order to avoid the algebraic loop problem. It is shown that this model, composed of a controlled voltage source in series with a resistance, can accurately represent four types of battery chemistries. The model's parameters can easily be extracted from the manufacturer's discharge curve, which allows for an easy use of the model. A method is described to extract the model's parameters and to approximate the internal resistance. The model is validated by superimposing the results with the manufacturer's discharge curves. Finally, the battery model is included in the SimPowerSystems (SPS) simulation software and is used in the Hybrid Electric Vehicle (HEV) demo. The results for the battery and for the DC-DC converter are analysed and they show that the model can accurately represent the general behaviour of the battery.
Article
A discussion is given of the derivation and application of an equation which gives an excellent description of a wide variety of cell and battery discharges. This equation gives the cell potential during discharge as a function of discharge time, current density, and certain other factors. It makes possible a complete description of cell discharge characteristics, using a minimum of experimental data and at the same time pinpointing experimental errors. It can also be used to describe cell charges, capacities, power evolution, and to predict capacities. A description is given of a simple numerical method for fitting this equation to a particular set of battery discharge curves. A graphical method is described for comparing characteristics of various types of cells.
Conference Paper
According to the measurement of the Ni-MH battery electrochemical impedance spectroscopy (EIS) and the relationship of the OCV-SOC , one basic equivalent-circuit model structure including hysteresis voltage for Ni-MH battery is presented, which can be applied for the estimation of battery state of charge (SOC). The model approach is evaluated by comparing the measured and the simulated battery voltage corresponding to several current profiles. Voltage hysteresis is shown to play a critical role in the Ni-MH battery model. Then the relationships among different physic variables and states variables of this model are expressed by the discrete-time state-space functions. Based on the Extended Kalman Filter, the Stare of Charge (SOC) for the Ni-MH battery model is estimated by using the date such as current and voltage which are real-time tested by the battery test instrument. Finally, the behavior of the algorithm in terms of accuracy is examined. Voltage hysteresis is shown to play a critical role, and this effect makes determining the SOC particularly difficult for Ni-MH batteries relative to other battery systems.
Article
A precise battery model is set up that gives the terminal voltage as a function of current and time. The model structure is based on a Randles' equivalent circuit. The non-linear dependence of all equivalent-circuit elements on current and frequency is taken into account. Model parameters are determined by impedance spectroscopy with various superimposed direct currents. Detailed measurements show that this battery model calculates the terminal voltage of lead/acid batteries with a tolerance of less than .2%.
Conference Paper
Effective amp hour (Ah) discharge capacity is an exponential function of discharge rate (current). Discharge and charge reversibly oxidize/reduce the active materials of the electrodes. Charging also produces water electrolysis. Terminal voltage shows hysteresis between charge and discharge. Tests of a previously reported voltage equalizing system showed the battery sensitivity to circuit impedance
He integrated in SPS the models of the electric battery and the fuel cell and he is author of two important demos of SPS: the multi-domain simulations of the power trains of an HEV
  • Technologie Supérieure
Technologie Supérieure, Montréal. He is in charge of developing new simulation models and solvers in the SimPowerSystems (SPS). He integrated in SPS the models of the electric battery and the fuel cell and he is author of two important demos of SPS: the multi-domain simulations of the power trains of an HEV and of an electric car based on a fuel cell.