Solar Energy 216 (2021) 551–563

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

HESS-based photovoltaic/batteries/supercapacitors: Energy management

strategy and DC bus voltage stabilization
Zineb Cabrane a, Jonghoon Kim a, *, Kisoo Yoo b, *, Mohammed Ouassaid c
a
Energy Storage and Conversion Laboratory, Department of Electrical Engineering, Chungnam National University, Daejeon 34134, Republic of Korea
b
Clean Energy Systems Laboratory, School of Mechanical Engineering, Yeungnam University, Gyeongsan 38541, Republic of Korea
c
Engineering for Smart and Sustainable Systems Research Centre, Mohammadia School of Engineers, Mohammed V University in Rabat, Rabat, Morocco

A R T I C L E I N F O A B S T R A C T

Keywords: The photovoltaic energy enables a variable power generation that is influenced by uncertain fluctuations caused
Hybrid energy storage system by the weather change (temperature and solar irradiation). Hence, the requirement for an energy storage system
Energy management strategy is essential to address this major issue. The use of only one energy storage element, such as battery, is insufficient.
DC-bus voltage stabilization
For this purpose, supercapacitors (SCs) can also be introduced as a power storage device. The combination of
Power converter
Photovoltaic system
batteries and SCs is a viable solution that requires an appropriate energy management strategy. The previous
studies are focused on the designing and modeling. In this study, a photovoltaic system with a hybrid energy
storage system (HESS) was developed by using batteries and supercapacitors. The development of an energy
management strategy requires energy distribution between two different storage mediums i.e. batteries and SCs.
Thus, a new dimension for a PI controller design is introduced to stabilize the direct current (DC) bus voltage and
command the buck-boost converters associated with batteries and SCs. The DC bus is controlled by voltage to
stabilize the DC bus voltage at the reference voltage 400 V. Two other PI controller are used to calculate the
reference current of batteries and SCs. The distribution of energy between batteries and SCs is based on a low
pass filter in order to eliminate the peak current on batteries and send it to SCs. The energy consumption of
batteries and SCs is controlled by the filter constant. The simulation results prove that the proposed sizing design
and system model provide good stabilization in the DC bus voltage. Furthermore, the augmented constant filter
reduces the state-of-charge (SOC) of the battery; hence the overall lifespans of batteries are increased. The
consumption of the SoCbat = 57.60% for τ = 2 s where the consumption of the SoCbat = 48.96% for τ = 16 s.

renewable sources is PV energy due to their low density of its prolifer­

1. Introduction
1.1. Motivation
Electricity production is growing rapidly, as it is strictly linked to the
population growth and economic development (Rezk et al., 2019).
However, the expansion of industrial tools and the growing need for
energy increase pollution and CO2 production. Renewable energy plays
an important role in ecofriendly energy production to reduce the waste
products created as a by-product from the fossil and fissile material
(Krishan and Suhag, 2020; Babazadeh-Dizaji et al., 2020). The most well
explored renewable sources are solar photovoltaic (PV) (Hernández-
Callejo et al., 2019), wind (Caporale et al., 2020), solar thermal (Kali­
dasan et al., 2020), hydro (Cheng et al., 2019), biomass (Aydin, 2019)
and ocean waves (Zhang et al., 2019). The most important among
ation in off-grid communities. This technology is not only cost effective
but also mature makes and easy to install (Jing et al., 2018; Jing et al.,
2017). The intermittent nature of the solar energy availability results in
drastic fluctuations in power generation using PV panels. Therefore, it is
necessary to develop a storage system that is able of controlling the
power fluctuations of the PV energy and load variations. In a standalone
PV system, an irregular charging pattern with battery storage system
shortens the battery lifespan. Batteries are characterised by a high en­
ergy density, but they suffer from a low power density, slow dynamic
response, and low charge/discharge rates (Mendiset al., 2014; Cabrane
et al., 2016). However, SCs produce high power density and low energy
capacity. SCs and batteries have complementary characteristics. In
hybrid energy storage system (HESS), they are combined to reduce the
size of the battery and increase its lifespan (Chong et al., 2016a, 2016b).

* Corresponding authors.
E-mail addresses: qwzxas@hanmail.net (J. Kim), kisooyoo@yu.ac.kr (K. Yoo).

https://doi.org/10.1016/j.solener.2021.01.048
Received 13 June 2020; Received in revised form 21 January 2021; Accepted 24 January 2021
Available online 8 February 2021
0038-092X/© 2021 International Solar Energy Society. Published by Elsevier Ltd. All rights reserved.
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

Nomenclature ISC SC pack current

vSC elementary supercapacitor voltage
Abbreviation iSC elementary supercapacitor current
DC direct current NP-SC number of parallel branches of SCs connections
SC supercapacitor NS-SC number of series branches of SCs connections
PV photovoltaic τ filter constant
EMS energy management strategy Td switching period
PI proportional integral Dpv duty cycle for PV
SOC state-of-charge Ipv photovoltaic current
MPPT maximum power point tracking Vpv photovoltaic voltage
HESS hybrid energy storage system Tpv control signal of the switch in boost converter
SoCbat state-of-charge of batteries Lpv boost inductance
SoCSC state-of-charge of SCs vLpv inductance voltage in boost converter
Vdc DC bus voltage iLpv inductance current in boost converter
Idc DC bus current ILpv Max maximum current in inductance in boost converter
Iph photocurrent in PV source model ILpv Min minimum current in inductance in boost converter
IS cell dark saturation current in PV source model vLsc inductance voltage of the buck- boost connected to the SCs
q electron charge (1.60217662 × 10− 19 C) VSC SCs voltage
k Boltzmann’s constant (1.3 × 10− 23 J/K) DSC duty cycle for SCs
ns-pv number of PV cells LSC buck-boost inductance connected to the SCs
TC cell’s working temperature in PV source model iLsc current of the inductance of the buck-boost connected to
A ideal factor the SCs
Rp shunt resistor in PV source model C0 initial SC value
RS series resistor in PV source model C1 main capacitance in SC model
Is,0 cell’s short-circuit current at a 25 ◦ C in PV source model Cv constant parameter in SC model
Tn cell’s reference temperature in PV source model Rload load resistant
Eg energy of the band gap Iscdc DC bus current of SCs
Iph light-generated current in PV source model Ibatdc DC bus current of batteries
Ir solar irradiation Ipvdc DC bus current of photovoltaic panels
In solar irradiation at TC tempertature Idcref. reference current of the DC bus
nb numbers of series cells of batteries Vdcref. reference voltage of the DC bus
Vbat battery voltage ISCref. reference current of supercapacitors
Ibat battery current Ibatref reference current of batteries
Ri constant resistance in battery model ωn pulsation
ΔTbat accumulator heating in battery ξ damping coefficient
Eb voltage source of battery τSC time constant for buck boost converter connected to the
Cbat battery capacitor SCs
C10 rated capacity in the battery τbat time constant for buck boost converter connected to the
νLbat inductance voltage of the buck-boost connected to the batteries
batteries Kpdc proportional gain for DC bus
iLbat inductance current of the buck-boost connected to the Kidc integral gain for DC bus
batteries KpSC proportional gain for SCs
Dbat duty cycle for batteries
KiSC integral gain for SCs
Cdc DC bus capacitor
USC SC pack voltage

1.2. Literatures 1.3. Contributions

The hybridization of batteries and SCs has been implemented suc­
cessfully in many applications, such as microgrid (Sinha and Bajpai,
2020), electric vehicle (Yang et al., 2020), uninterruptible power supply
(Lahyani et al., 2013), solar vehicle (Cabrane et al., 2020) and pumping
system (Das and Mandal, 2018). The energy management was devel­
oped in many papers. M. Aryanezhad presents the management and
coordination of LTC, SVR for the high PV penetration in power distri­
bution system for voltage regulation and power loss minimization
(Aryanezhad, 2018). M. Aryanezhad presents the management and co­
ordination charging of smart park and V2G strategy based on Monte
Carlo algorithm (Aryanezhad et al., 2014). Y. Han gives the hierarchical
energy management for PV, hydrogen, battery for an island DC micro­
grid (Han et al., 2019).
The innovative contributions presented in this article are as follows:
• A comparison of different topologies used to connect batteries and
SCs to the load. This comparison is presented by different schemas
and gives the advantage and disadvantage of each topology.
• A design of HESS using batteries and SCs for PV energy storage.
• A new method of controlling by voltage of the DC bus and controlling
by current the buck-boost converter is introduced. The advantage of
providing a control strategy is to stabilize the DC bus voltage.
• An energy management strategy (EMS) is proposed to distribute
energy between batteries and SCs.
1.4. Organization

This work is organized as follows: After introduction in Section 1, the

552
Z. Cabrane et al.
different topologies of hybrid system are given in Section 2. The study
and modeling of global system is presented in section 3. Section 4 is
reserved to the simulation results and validation. Concluding remarks
are given in the final section. The schematic overview of the proposed
approach is given in Fig. 1.
2. Comparison of different topologies used for HESS based on
batteries and SCs
Batteries and SCs can be combined to achieve hybridization that can
be used in different renewable energy systems, especially PV systems.
Consequently, there is a multitude of possibilities for exchanging power
between these two components. The choice of the most appropriate
configuration is relative to the application type and the installed power.
The most used topologies in the HESS are described below (Jing et al.,
2018).
2.1. HESS using hybrid passive parallel configuration
A hybrid passive parallel configuration is considered a feasible so­
lution in combing batteries, SCs and the DC bus. This configuration is
given in Fig. 2. In hybrid passive parallel configuration, the load, SCs
and the batteries are connected in parallel because, in this configuration
SC starts to behave like a low pass filter. The summary of the pros & cons
of the parallel passive configuration is given in Table 1 (Kouchachvili
et al., 2018).
2.2. HESS using semi-active configuration using SCs
The semi active configuration by SCs is shown in Fig. 3. This HESS
Fig. 1. Schematic overview of the proposed approach.
553
Solar Energy 216 (2021) 551–563
configuration consists of a single traditional buck-boost converter, SCs
and batteries in parallel connection. A controlled power system is
required for the semi-active configuration. However, the buck-boost
converter should be oversized to manage the power of the SCs (Chong
et al., 2018; Shen and Khaligh, 2015). In this configuration, the voltage
of SCs can be used in a wider range. The advantages and disadvantages
of the semi-active configuration with SCs are presented in Table 2.
2.3. HESS using semi-active configuration using batteries
The semi-active configuration by batteries is illustrated in Fig. 4. The
configuration requires a single buck-boost incorporated with batteries in
parallel. The battery voltage can be higher and lower than the SCs
voltage. Here, the SCs behave like a low-pass filter (Zhao et al., 2015).
The advantages and disadvantages of the semi-active battery configu­
ration are presented in Table 3.
2.4. HESS using multiple converters configuration
The multiple converter configuration is illustrated in Fig. 5. The
configuration requires two buck-boost converters with batteries and SCs
in parallel separately. In this scenario, the overall voltage of the batteries
and SCs can be lower than the DC bus voltage. The advantages and
disadvantages of the configuration by multiple converters are presented
in Table 4 (Vargas et al., 2019; Roy et al., 2019).
3. Modeling and control strategy of the HESS
The system under study is composed of PV panels, batteries and SCs.
After the comparison of different configurations, the topology chosen is
the configuration with multiple converters. The modeling of different
converters is shown in detail. The sizing of the proportional–integral (PI)
controllers is also detailed.
3.1. Modeling of PV panels
The electrical circuit of the chosen PV model is presented in Fig. 6. It
is by one diode and a photocurrent. The leakage current are expressing
with a shunt resistor Rp. An internal resistance represented with a series
resistor Rs that represents the current flow (Rekioua et al., 2014). The PV
current Ipv is given by (Poshtkouhi et al., 2012).
( ( ) )
q(Vpv +Ipv Rs ) I R +V
Ipv = Iph − Is exp ns− pv AkTc − 1 − pv Rs p pv (1)
The saturation current Is is presented by the following equation:
( )3 [ ( )]
qE
Is = Is,0 TTn exp Akg T1n − T1 (2)where Is,0 is the short-circuit cur­
rent calculated at a 25 ◦ C and 1 kW/m2. The photocurrent Iph is
expressed as (3) where ΔT = T – Tn.
( )
Iph = Iph,n + KI ΔT GGn (3)
The power-voltage curves of PV panels with different values of solar
irradiation is presented in Fig. 7.
3.2. Modeling of the boost converter associated with PV panels
A boost converter is associated with PV panels for DC-DC conversion.
The current Ipv and voltage Vpv are considered to be state variables for
the converter, and Dpv is the duty cycle controlling the converter. The
maximum power point tracking (MPPT) control algorithm gives Dpv
value. It is between 0 and 1. To simplify the modeling of the converter, it
is assumed to be lossless. The control signal of the switch Tpv is illus­
trated in Fig. 8(a). It is a rectangular signal of frequency f, whose
duration is adjusted by the duty cycle Dpv. The duty cycle is defined as
the ratio between the conduction duration of the switch Tpv and the
switching period Td and is represented by the following equation.
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

Fig. 2. Structure of a hybrid passive parallel configuration.

is reverse biased. The equivalent circuit is shown in Fig. 8(b). The

Table 1
voltage applied across the inductance is given by the following differ­
Pros and cons of a hybrid passive parallel configuration.
ential equation:
Pros • Implementation is easy, and no complicated control device is necessary. di
• The reliability is greater due to the reduction in the number of elements at
vLpv = Vpv = Lpv dtLpv (6)
risk of failure. The formula that expresses the evolution of the current through the
• There is no conversion loss. inductor is obtained by solving the differential Equation (6). It can be
Cons • The distribution of power between the SC and the batteries is not
formulated as follows:
controlled but depends on the internal resistance of batteries and SCs. Vpv
• The voltage of the DC bus is not constant and changed continuously with iLpv = Lpv t +ILpv Min (7)
respect to battery voltage.

• Freewheeling phase (Fig. 8(c))

TON
Dpv = (4) When the switch is open, at time t = DpvTd, during a duration of Td (1-
Td
Dpv), the inductive current cannot change instantaneously. The voltage
Td = TON + TOFF (5)
Table 2
where TOFF corresponds to the blocking duration of the switch (TOFF =
( ) Pros and cons of a semi-active configuration using SCs.
1 − Dpv Td ) and TON corresponds to the conduction time of the switch
Pros • Ability to fully and more efficiently use the energy of SCs.
(TON = Dpv Td ). Depending on the state of the two switches Tpv and Dpv,
• The voltage of the DC bus is relatively constant, and it changes with respect
two phases of operation can be distinguished, namely the active and the to the SOC of the batteries because of direct connection of the batteries and
freewheel phases, which are shown in Fig. 8. the DC bus.
Cons • The voltage regulation of the DC bus is difficult, because, in this scenario,
• Active phase the batteries are connected to the bus directly.
• Voltage oversizing may be necessary. For this, the voltage level must be the
same as that of the DC bus, also increasing the risks of load imbalance.
When the switch is closed, at t = 0, for a duration of DpvTd, the diode

Fig. 3. Structure of a semi-active configuration using SCs.

554
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

Fig. 4. Structure of a semi-active configuration using batteries.

diLSC
Table 3 vLSC = VSC − Vdc = LSC dt (13)
diLSC VSC − Vdc
Pros and cons of a semi-active configuration using batteries. = (14)
dt LSC
Pros • One converter that transmits the average power of the load is used. The electrical equation of the boost converter, which links the two
• The sizing and mass cost are less than the configuration using multiple Equations (11) and (13), is written as follows:
converters. diLSC
Cons • The regulation of the DC bus voltage is poor because of the direct LSC dt = VSC − (1 − DSC )Vdc (15)
connection of the SCs to the DC bus.
• The voltage of SCs varies when they supply or receive energy.
3.4. Control and stabilization of the DC bus voltage

across the inductor is expressed as follows: The DC bus voltage is controlled according to the principle described
di in Fig. 10. The reference current of the DC bus Idcref is calculated by a PI
vLpv = Vpv − Vdc = Lpv dtLpv
(8)
controller, which maintains the DC bus voltage Vdc at the reference
The expression of the evolution of the current passing through the
voltage Vdcref = 400 V (Cabrane et al., 2017). The EMS provides the
inductance is obtained by solving the differential Equation (4). It can be
reference currents (Ibatref and ISCref) to control the buck-boost converter
written as follows:
V − V ( ) connected to the batteries and supercapacitors. These currents are
iLpv = pvLpv dc t − Dpv Td +ILpv Max (9) responsible for stabilizing the voltage of the DC bus irrespective to the
The time average of Equations (6) and (8) gives an average model of uncertain power from the PV source and the load. As shown in Equation
this converter. It is written as follows: (16), the Idcref is equal to the sum of ISCref and Ibatref.
Lpv
diLpv
= Vpv − (1 − Dpv )Vdc (10) Idcref = ISCref +Ibatref (16)
dt
The peaks in the currents can be eliminated by introducing a low-
pass filter in the EMS, because it gives an exponential form to Ibatref.
3.3. Modeling of the buck-boost converter associated with SCs ISCref and Ibatref are written as follows:
( ( ))
To connect the batteries or SCs to the DC bus, a parallel converter, Ibatref = Idcref 1 − exp − τt (17)
( )
reversible in current (step-down) is required to associate the SCs or the ISCref = Idcref exp − τt (18)where τ is the filter constant time.
batteries with the DC bus (Fig. 9).

• Active phase (Fig. 9(a)): 3.5. Control strategy used to stabilize the DC bus voltage

When the switch is closed, the voltage is defined by the following To manage the energy transfers between the different components,
equations: the DC bus is controlled with voltage by a PI controller. In this case, the
buck-boost converter associated with the SCs is responsible for calcu­
vLsc = Vsc = Lsc didtLsc (11)
diLsc
lating the parameters of this corrector. A resistance rL is added to this
dt = VLscsc (12) model in parallel with the load represent the different losses. The
equations for the proposed converter are given below:
• Freewheeling phase (Fig. 9(b)): Cdc dVdtdc = iLSC (1 − DSC ) − Vdc
− idc (19)
rL
di
When the switch is open, the current in the inductor cannot change LSC dtLSC
= VSC − (1 − DSC )Vdc (20)
instantaneously. This results in a direct polarized of diode. In this con­ Therefore αSC = 1 − DSC . Then,
dition, the voltage across the inductor is written as follows: Cdc dVdtdc = iLSC αSC − Vdc
rL − idc (21)

555
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

Fig. 5. Structure of a multiple converters configuration (Carcia et al., 2012).

Table 4
Pros and cons of multiple converters configuration.
Pros • The power of the batteries and the power of the SCs can be separately
controlled according on their SOC and power requirements.
Cons • The number of components is greater.
• The cost is higher.

Fig. 6. Electrical circuit of PV panels.

Fig. 7. Power-voltage curves of PV panels with different values of solar
irradiation.
di
LSC dtLSC = VSC − αSC Vdc (22)
From the above equations, one obtains the following dynamic is deduced.
equation: V
FTx(s) == Idcdc(s) VSC rL
(27)
(s) = rL CSC s+1
Cdc dVdtdc = iLSC VVSC − Vdc
− idc (23) 2
dc rL The transfer function of the PI controller is given by the following
2
To have a linear function, X = Vdc is replaced: equation.
dX
dt = 2Vdc dVdtdc (24) Cdc (s) = Kpdc +
Kidc
(28)which gives:
s
It is deduced that: ( )
Ki
dVdc Kpdc rL VSC s+Kpdc
dt = 2V1dc dX
dt (25)and the following relation is obtained: dc
FTBF(s) = ( ) ( ) (29)which gives:
Cdc dX X
dt = 2iLSC VSC − 2 rL − 2idc Vdc (26) Kpdc rL VSC
Ki
s+Kpdc +s
rL CSC
s+1
2
The transfer function of the DC bus expressed in the Laplace domain
dc

556
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

Fig. 8. Average model of a boost converter: (a) Switch control signal Tpv, (b) Closed switch mode, (c) Opened switch mode.

Fig. 9. Average model of a buck-boost converter: (a) Closed switch mode, (b) Opened switch mode.

Fig. 10. Schematic diagram of the hybrid energy storage system (HESS) control strategy.

( )
Ki By identifying the denominator with that of the canonical form, we
Kpdc rL VSC s+Kpdc
dc obtain
FTBF(s) = r (30) ⎧
L CSC 2
2 s +
Ki
(Kpc rL VSC +1)s+Kpdc rL VSC Kpdcdc 2 2 2 ( )
⎪ ωn = C VSC Kidc
⎪
⎪ 2ξωn = Kpdc rL VSC + 1
The closed-loop transfer function is written as:
⎨ SC rL CSC
( ) where (32)
Ki
⎪ K = ωn CSC
⎪
⎪ 2 ξωn rL CSC − 1
2
CSC Kpdc VSC s+Kpdc ⎩ K pdc =
dc idc
2VSC rL VSC
CLTF(s) = s2 + 2 2 (31)
rL CSC (Kpdc rL VSC +1)s+CSC VSC Kidc

557
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

3.6. Control strategy used for buck-boost converter 4, 8, 10, 16 s]. These tests were conducted with the following conditions:
- Temperature: 25 ◦ C
For the control of the buck-boost converter, the following equation is - Initial state of charge of batteries: SoCbat = 80%
- Initial state of charge of SCs: SoCSC = 57%
used:
- Variation of the solar current
VSC = LSC iLSC (s)s +RLoad iLSC (s) +(1 − DSC (s) )Vdc (33) - Variation of the load current
Thus, the equation that relates the duty cycle DSC and the current ISC
is obtained as follows:
Vdc
ISC (s)
(34)
RLoad
DSC (s) = L
1+R SC
Load
s 4.1. Simulation results of HESS with batteries and SCs
The transfer function of the corrector is as follows:
( )
Kp The variation of the PV current is given in Fig. 12(a). This is follows
KiSC 1+ K SC s
iSC
the same change of the solar irradiation. It is represented by [100, 500,
GSC (s) = s (35)
300, 250A]. The proposed load current is given in Fig. 12(b) that vari­
In order to simplify the transfer function of the system, a pole/zero
KpSC LSC
ance between 170A and 1320A. This variation is chosen in order to be
and imposing compensation are assumed: KiSC = RLoad . The new closed- less and more than PV current. For this change of load current and PV
loop transfer function becomes current, batteries and SCs can have the charge and discharge status. The
CLTF(s) = 1
RLoad where τSC = VRdcLoad
Ki (36)
DC bus voltage is given in Fig. 12(c), which considered constant at 400 V
1+V s
with some fluctuations for changes of the solar irradiation and load
SC
dc KiSC

Then, current. The simulation results of the test with τ = 2 s is shown in Fig. 13.
KiSC = τRSCLoad LSC
Vdc , KpSC = τSC Vdc (37) This simulation test is represented with the same variation of the solar
The sizing and modeling of the global HESS is given in Fig. 11. irradiation and load current. The variation of the batteries current is
represented in Fig. 13(a). The variation of SCs current is shown in Fig. 13
4. Simulation results and validation (b). The simulation results of the test with τ = 4 s is given shown in
Fig. 14. This simulation test is given with the same variation of the solar
To verify the performance of the proposed EMS and control system, irradiation and load current. The fluctuations of the batteries current
simulation results were obtained using MATLAB/Simulink software. and SCs current are represented in Fig. 14(a) and 14(b), respectively.
Five simulation tests are compared with different filter constant τ = [2, The simulation results of the test with τ = 8 s is given shown in Fig. 15.
This simulation test is given with the same variation of the solar irra­
diation and load current. The fluctuations of the batteries current and
SCs current are represented in Fig. 15(a) and 15(b), respectively. The

Fig. 11. Schematic diagram of the HESS control strategy after sizing.

558
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

Fig. 12. Analytic simulation results for the PV energy storage by using batteries and SCs: (a) Photovoltaic current (W/m2), (b) Load current (A), (c) DC bus
voltage (V).

Fig. 13. Analytic simulation results with τ = 2 s: (a) Batteries current (A), (b) SCs current (A).

Fig. 14. Analytic simulation results with τ = 4 s: (a) Batteries current (A), (b) SCs current (A).

559
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

Fig. 15. Analytic simulation results with τ = 8 s: (a) Batteries current (A), (b) SCs current (A).

Fig. 16. Analytic simulation results with τ = 10 s: (a) Batteries current (A), (b) SCs current (A).

simulation results of the test with τ = 10 s is given shown in Fig. 16.
These simulation tests are given with the same variation of the solar
irradiation and load current. The fluctuations of the batteries current
and SCs current are represented in Fig. 16(a) and 16(b), respectively.
The simulation results of the test with τ = 16 s are given in Fig. 17. The
variation of the batteries current is given in Fig. 17(a). The variation of
the SCs current is given in Fig. 17(b). Batteries react slowly to the de­
mands when the SCs eliminate the peak current on batteries. In this case
(τ = 16 s), SCs provides more power than batteries.
4.2. Comparison of simulations with different values of τ
A comparison of five simulation tests with different values of the
filter constant τ = [2 4 8 10 16 s] is given in Fig. 18. A comparison of
SOC of SCs for different value of τ is represented in Fig. 18(a). The SoCSC
decreases by the increasing of the filter constant τ. The comparison of
the SoCbat for different value of τ is given in Fig. 18(b). The SoC of
batteries increase when the filter constant increases. A comparison of
SCs current for τ = 2 s and τ = 16 s is given in Fig. 18(c). It the simulation
of the system with τ = 2 s, the SCs eliminate just peak current. However,
for the simulation of system with τ = 16 s, the SCs eliminate the peak
current and insure a long charge/discharge current. The simulation re­
sults for different times, t = [250, 300, 400, 500, 600 s] and different
filter constant τ = [2, 4, 6, 8, 10, 12, 16 s] in given in Fig. 19. The SOC of
batteries where they increase with the increase of the filter constant is
given in Fig. 19(a). The consumption of batteries decreases with the
increase of the filter constant, as shown in Fig. 19(b). A tree-dimensional
representation of SOC of batteries and consumption are provided by
Fig. 19(c) and (d), respectively.
The consumption of energy by batteries for 250 s and the estimation

Fig. 17. Analytic simulation results with τ = 16 s: (a) Batteries current (A), (b) SCs current (A).

560
Z. Cabrane et al. Solar Energy 216 (2021) 551–563

Fig. 18. Comparison of analytic simulation results with different value of the constant filter τ: (a) SOC of SCs (%), (b) SOC of batteries (%), (c) SCs current (A).

(a) 77 (b) 9
t=250s t=250s
Batteries Consumption (%)

76 t=300s 8 t=300s
75 t=400s t=400s
7
t=500s t=500s
SoCbat (%)

74 t=600s 6 t=600s
73
5
72
4
71
3
2 4 6 8 10 12 14 16 18 20 22 24 2 4 6 8 10 12 14 16 18 20 22 24
(s) (s)

(c) (d)
Battries Consumption (%)

80 10
SoCbat (%)

75 8

70 6 600
0 500
2 600 4
4 2 4 400
6 500 6 8
8 10 300 Time (s)
(s) 1012 400 (s) 12 14
16 200
14 300 Time (s)
16
Fig. 19. Analytic simulation results with different time and filter constant: (a) SOC of batteries, (b) Batteries consumption, (c) 3d representation of SOC of batteries,
(d) 3d representation of batteries consumption.

561
Z. Cabrane et al.
Table 5
Consumption and estimation of the SoCbat for one hour.
SoCSC for 250 s (%) SoCbat for 250 s (%)
τ=2s 58 76.0
τ=4s 55 76.1
τ=8s 49 76.2
τ = 10 s 48 76.3
τ = 16 s 50 76.6
of the SoCbat for one hour is given in Table 5. The consumption of bat­
teries SOC reduce by the increasing of the filter constant. The con­
sumption of the SoCbat = 57.60% for τ = 2 s where the consumption of
the SoCbat = 48.96% for τ = 16 s.
The proposed sizing of PI controller gave good results and stabilized
the DC bus voltage at 400 V, as shown in Fig. 9(c). The proposed EMS
distributed the energy correctly between batteries and SCs by consid­
ering the fluctuation of the solar irradiation and the load variation.
5. Conclusion
The implementation of a new control system for the PV energy
storage that uses batteries and SCs was evaluated. This control system is
based on PI controllers. A control strategies using voltage command was
introduced to stabilize the DC bus voltage. The control of the converters
associated with batteries and SCs is based on a current control. An EMS
design was proposed to incorporate batteries, SCs, PV energy and load.
This EMS delivers also the reference current of batteries and SCs. To
prove the effectiveness of the proposed control strategy, simulation tests
was conducted with different filter constants τ = [2, 4, 8, 10, 16 s].
These simulation tests proved that the proposed EMS is able to distribute
energy between different devices. Furthermore, a comparison of the SOC
of batteries with different values of the filter constant demonstrated that
the integration of SCs reduced the consumption of batteries and elimi­
nate the peak current. Indeed, the consumption of batteries SOC reduce
by the increasing of the filter constant. The consumption of the SoCbat =
57.60% for τ = 2 s where the consumption of the SoCbat = 48.96% for τ
= 16 s.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
This research was supported by a grant (20TLRP-C135446-04,
Development of Hybrid Electric Vehicle Conversion Kit for Diesel De­
livery Trucks and its Commercialization for Parcel Services) from the
Transportation & Logistics Research Program (TLRP) funded by the
Ministry of Land, Infrastructure and Transportation of the Korean gov­
ernment and Korea Institute of Energy Technology Evaluation and
Planning (KETEP) and the Ministry of Trade, Industry & Energy
(MOTIE) of the Republic of Korea (No. 20182410105280).
Appendix A. Supplementary material
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.solener.2021.01.048.
References
Aryanezhad, M., 2018. Management and coordination of LTC, SVR, shunt capacitor and
energy storage with high PV penetration in power distribution system for voltage
Solar Energy 216 (2021) 551–563
Consumption of SoCbat for 250 s (%) Estimation of the consumption of SoCbat for 1 h (%)
4.0 (4.0 × 3600)/250 = 57.60
3.9 (3.9 × 3600)/250 = 56.16
3.8 (3.8 × 3600)/250 = 54.73
3.7 (3.7 × 3600)/250 = 53.28
3.4 (3.4 × 3600)/250 = 48.96
regulation and power loss minimization. Int. J. Electr. Power Energy Syst. 100,
178–192.
Aryanezhad, M., Ostadaghaee, E., Joorabian, M., 2014. Management and coordination
charging of smart park and V2G strategy based on Monte Carlo algorithm. 2014
Smart Grid Conference (SGC).
Aydin, M., 2019. The effect of biomass energy consumption on economic growth in
BRICS countries: A country-specific panel data analysis. Renewable Energy 138,
620–627.
Babazadeh-Dizaji, R., Hamzeh, M., Hekmati, A., 2020. A frequency-based economical-
sharing strategy for low-voltage DC microgrids. Int. J. Electr. Power Energy Syst.
118, 105822.
Cabrane, Z., Ouassaid, M., Maarouf, M., 2016. Analysis and evaluation of battery-
supercapacitor hybrid energy storage system for photovoltaic installation. Int. J.
Hydrogen Energy 41, 20897–20907.
Cabrane, Z., Batool, D., Kim, J., Yoo, K., 2020. Design and simulation studies of battery-
supercapacitor hybrid energy storage system for improved performances of traction
system of solar vehicle. J. Storage Mater. 32, 101943.
Cabrane, Z., Ouassaid, M., Maaroufi, M., 2017. Battery and supercapacitor for
photovoltaic energy storage: a fuzzy logic management. IET Renew. Power Gener.
11, 1157–1165.
Caporale, D., Sangiorgio, V., modio, A., Lucia, C. D., 2020. Multi-criteria and focus group
analysis for social acceptance of wind energy. Energy Policy 140, 111387.
Cheng, C., Blakers, A., Stocks, M., Lu, B., 2019. Pumped hydro energy storage and 100 %
renewable electricity for East Asia. Global Energy Interconnection 2, 386–392.
Chong, L.W., Wong, Y.W., Rajkumar, R.K., Isa, D., 2016a. An optimal control strategy for
standalone PV system with Battery-Supercapacitor Hybrid Energy Storage System.
J. Power Sources 331, 553–565.
Chong, L.W., Wong, Y.W., Rajkumar, R.K., Isa, D., 2016b. Hybrid energy storage systems
and control strategies for stand-alone renewable energy power systems. Renew.
Sustain. Energy Rev. 66, 74–189.
Chong, L.W., Wong, Y.W., Rajkumar, R.K., Isa, D., 2018. An adaptive learning control
strategy for standalone PV system with battery-supercapacitor hybrid energy storage
system. J. Power Sources 394, 35–49.
Das, M., Mandal, R. A., 2018. Comparative performance analysis of direct, with battery,
supercapacitor, and battery-supercapacitor enabled photovoltaic water pumping
systems using centrifugal pump, Sol. Energy 171, 302-309.
García, P., Torreglosa, J.P., Fernández, L.M., Jurado, F., 2012. Viability study of a FC-
battery-SC tramway controlled by equivalent consumption minimization strategy.
International Journal of Hydrogen Energy. (370), 9368–9382.
Han, Y., Zhang, G., Li, Q., You, Z., Chen, W., Liu, H., 2019. Hierarchical energy
management for PV/hydrogen/battery island DC microgrid. Int. J. Hydrogen Energy
44, 5507–5516.
Hernández-Callejo, L., Gallardo-Saavedra, S., Alonso-Gómez, V., 2019. A review of
photovoltaic systems: Design, operation and maintenance. Sol. Energy 188,
426–440.
Jing, W., Lai, C.H., Wong, W.S.H., Wong, M.L.D., 2018. A comprehensive study of
battery-supercapacitor hybrid energy storage system for standalone PV power
system in rural electrification. Appl. Energy 224, 340–356.
Jing, W., Lai, C.H., Wong, W.S.H., Wong, M.L.D., 2017. Dynamic power allocation of
battery-supercapacitor hybrid energy storage for standalone PV microgrid
applications. Sustainable Energy Technol. Assess. 22, 55–64.
Kalidasan, B., Pandey, A.K., Shahabuddin, S., Samykano, M.,
Thirugnanasambandam, M., Saidur, R., 2020. Phase change materials integrated
solar thermal energy systems: Global trends and current practices in experimental
approaches. J. Storage Mater. 27, 101118.
Krishan, O., Suhag, S., 2020. Grid-independent PV system hybridization with fuel cell-
battery/supercapacitor: Optimum sizing and comparative techno-economic analysis.
Sustain. Energy Technol. Assess. 37, 100625.
Kouchachvili, L., Yaïci, W., Entchev, E., 2018. Hybrid battery/supercapacitor energy
storage system for the electric vehicles. J. Power Sources 374, 237–248.
Lahyani, A., Venet, P. Guermazi, A. Troudi. A., 2013. Battery/ supercapacitors
combination in uninterruptible power supply (UPS). IEEE Trans. Power Electron. 28,
1509-1522.
Mendis, N., Muttaqi, K.M., Perera, S., 2014. Management of low- and high-frequency
power components in demand-generation fluctuations of a DFIG-based wind-
dominated RAPS system usinghybrid energy storage. IEEE Trans. Ind. Appl. 50,
2258–2268.
Rezk, H., A. Nassef, M., Abdelkareem, M. A., Alami, A. H., Fathy, A., 2019. Comparison
among various energy management strategies for reducing hydrogen consumption in
a hybrid fuel cell/supercapacitor/battery system. Int. J. Hydrogen Energy.
Rekioua, D., Bensmail, S., Bettar, N., 2014. Development of hybrid photovoltaic-fuel cell
system for stand-alone application. Int. J. Hydrogen 39, 1604–01611.
562
Z. Cabrane et al.
Roy, P.K.S., Karayaka, H.B., Yan, Y., Alqudah, Y., 2019. Investigations into best cost
battery-supercapacitor hybrid energy storage system for a utility scale PV array.
J. Storage Mater. 22, 50–59.
Sinha, S., Bajpai, P., 2020. Power management of hybrid energy storage system in a
standalone DC microgrid. J. Storage Mater. 30, 101523.
Shen, J., Khaligh, A., 2015. A Supervisory Energy Management Control Strategy in a
Battery/ Ultracapacitor Hybrid Energy Storage System. IEEE Trans. Transp.
Electrificat. 1, 223–231.
Vargas, U., Lazaroiu, G.C., Tironi, E., Ramirez, A., 2019. Harmonic modeling and
simulation of a stand-alone photovoltaic-battery supercapacitor hybrid system.
Electrical Power Energy Syst. 105, 70–78.
Yang, B., Wang, J., Zhang, X., Wang, J., Shu, H., Li, S., He, T., Lan, C., Yu, T., 2020.
Applications of battery/supercapacitor hybrid energy storage systems for electric
563
Solar Energy 216 (2021) 551–563
vehicles using perturbation observer based robust control. J. Power Sources 448,
227444.
Zhang, D., Shi, J., Si, Y., Li, T., 2019. Multi-gratingtriboelectricnano generator for
harvesting low-frequency ocean wave energy. Nano Energy 61, 132–140.
Zhao, D., Yin, H., Yang, Z., Ma, C., 2015. Equivalent Series Resistance-Based Energy Loss
Analysis of a Battery Semiactive Hybrid Energy Storage System. IEEE Trans. Energy
Convers. 30, 1081–1091.
Poshtkouhi, S., Palaniappan, V., Fard, M., Trescases, O., 2012. A general approach for
quantifying the benefit of distributed power electronics for fine grained MPPT in
photovoltaic applications using 3-D modeling. IEEE Trans Power Electron 27,
4656–4666.