An Improved Approach of Control for a Battery

Charger Based Forward Converter and SEPIC

Abdelilah Hassoune*, Mohamed Khafallah, Tarik Bouragba

Abdelouahed Mesbahi, EIGSI Casablanca
Laboratory of Energy & Electrical Systems (LESE) Casablanca, Morocco
Hassan II University, ENSEM
Casablanca, Morocco
[email protected]

Abstract—This work presents an improved approach of control topology is based on a DC bus medium voltage, connects all
for charging a lithium-ion battery with two different topologies of renewable energy sources via specific converters and all kind
chargers i.e., forward converter and single ended primary of loads such as, EV batteries and residential AC loads,
inductor converter (SEPIC). In order to fulfill the customer besides, the grid is also combined in this platform [7].
power requirements in terms of accuracy and rapidity, the
electrical system gathered topology/control is testing all the The number one priority of an EV customer is to charge his
performance constraints such as, the instability of the power vehicle battery through a rapid charger. In the meanwhile, the
system and how the energy conversion devices would react EVCS infrastructure must equipped with the appropriate
efficiently to any kind of perturbations. Simulation results show technology of converters in order to fulfill the settled rate of
the effectiveness of the proposed topologies operated under a charging power from each connected EV [8]. According to [9],
management algorithm of voltage/peak current mode control. superchargers of 350 kW are designed to feed the EV battery
within 8 minutes, to support this great amount of energy flow
Keywords-component; Forward converter; SEPIC; in short time, the technology of the battery must handle the
management algorithm, lithium-ion; battery charger. high rate of injected current to preserve its lifetime from the
overheating phenomena and other inconvenient concerns [10].
I. INTRODUCTION Furthermore, the supply system of the EVCS is now under a
In recent years, the increased emission of greenhouse gases massive power demand, especially when several vehicles are
and the decreased level of fossil fuels reserve have caused plugged in the same time. The converters as well must be
limited sources of energy in electrical systems based petroleum designed by which their components could bear the burden
products [1]. However, the adoption of distributed power from power/frequency. Moreover, SiC and GaN technologies
generation has received high attention due to its various have proven to be more effective over silicon by which a better
benefits, as the improved accuracy, stability and decreasing the switching performance is achieved under great power demand
energy losses within the conversion energy devices (CED). [11]. The safety measurements are primordial in such precise
Another advantage of the CED is presented during the process operation modes, where DC charging has to be fitted with
of establishing a hybrid energy source, which has different detached blocks to set up an isolation of outputs from inputs.
platforms and methods of control [2]. This work describes a comparison between two different
The integration of renewable energy sources in the power topologies of dc/dc converters i.e., forward converter and
systems based grid in a standalone mode has become the topic single ended primary inductor converter (SEPIC) running in
of several researches [3]. The intermittent in solar irradiance the buck/boost mode. The accuracy and the stability of control
and wind speed are recognized as a key element where accurate would be tested under the fast variations of the input current
CED would configure the required control to obtain quite a ripples [12]. The electrical system chargers are often being the
high level of efficiency [4]. A suitable power converter is used most critical phases where isolated converters are required to
in multiple frameworks e.g., extract the maximum power from get separate grounds due to the high frequency transformer
PV cell and from variable speed of wind turbine blades. Apart providing that case feasibility. Another outcome of an isolated
the control techniques, the architecture of power converters is converter is its outlet that can be set to be either positive or
often considered important to be fitted with extra features i.e., negative. The forward converter along with flyback converter
high power yield and stability of voltage/current rates in both are known as low power isolated CED, they are representing an
sides, input and output of electrical devices [5]. improved version of the classic dc/dc converters where many
enhancements are obtained e.g., increased noise immunity [13].
The electric vehicle charging station (EVCS) presents one In this project, a 50W prototype of a charger and lithium-ion
of the main applications of the charger, it contains almost all battery, are set to simulate each topology with voltage mode
the energy conversion types i.e., dc/dc, dc/ac, ac/dc and ac/ac. control (VMC) and current mode control (CMC). A power
Recently, multiple architectures of EVCS are discussed in the management algorithm is expressed in order to integrate all
literature to treat all the energy constraints [6]. The efficient these criteria in one efficient platform [14].

 
 
 The paper is divided as follows. Section 1 introduces the allowing the dissipation of the output current from Lch to the
context of this work. The topology and the description of the load while the magnetizing current through D3 goes down to
adopted dc/dc converters are provided in Section 2. Design and zero [17].
simulation results in MATLAB/Simulink of the forward
converter and the SEPIC operated under VMC and peak CMC B. Single Ended Primary Inductor Converter (SEPIC)
are performed in Section 3. Conclusions of this paper are
The SEPIC is a dc/dc converter, the electrical voltage at its
presented in Section 4.
output can be less than, greater than or equal to that at its
input. The configuration of SEPIC generally involves a boost
II. DESCRIPTION OF THE PROPOSED CONVERTERS converter followed by a buck-boost converter. The output of
the SEPIC control is performed by an accurate duty cycle for
A. Forward converter the single transistor. Compared to the buck/boost converter,
Derived from the buck diagram, the forward converter SEPIC has the same voltage polarity as the input that is its
based single transistor composed by a transformer and that main advantage [18].
refers to galvanic isolation, as well as increase voltage or
decrease it, which makes it a suitable choice for specific The Fig. 2 shows the scheme of SEPIC. When the power
applications requiring both [15]. Moreover, the technology of switch is turned on, the voltage set-up and when the switch is
the single active switch must take into account the turned off the voltage drops. During the steady-state process,
power/frequency dilemma forced by the specifications of the the ripple voltage is neglected.
connected load. Fig. 1 depicts the scheme of the single
transistor forward converter, charging a capacitive load via
lithium-ion battery.

Figure 2. Schematic of SEPIC

In order to understand the operating phases of the

converter, it is necessary to analyze the diagram taking in
consideration the two states of the transistor (S1) during
continuous conduction mode (CCM) i.e., activated during the
first phase and off at the second. The CCM means the current
Figure 1. Schematic of a single transistor forward converter through the inductor L1 never goes down to zero. However,
during the steady-state operation, the average voltage across
The transferring energy during the conduction time of the capacitor C1 is equal to the input DC link voltage (VIN). Fig. 3
switching phase is not stored. Unlike buck/boost converter, this depicts the SEPIC operating in CCM, where the voltage across
diagram contains transformer instead of inductors which D1 is neglected [19].
reduces the air gap due to the high magnetizing inductance of
the transformer [16]. The output voltage of the forward (a)
converter is expressed in Eq. (1).

(1)

Where, VBAT is the battery voltage, VIN is the input voltage,

NP and NS are the number of winding turns of a transformer,
primary and secondary respectively, DC is the pulse width
modulation duty cycle of the transistor. (b)
The typical diagram operation of the forward converter is
described as follows: When the control drive is set to the
switch (S), the current of the secondary bloc will flow through
D1, Lch, and into the battery. This operation mode is due to
transformer effect. In the meanwhile, the magnetizing current
starts to accumulate in the transformer primary. When S is off,
then the magnetizing current would be dissipated in the
primary windings. The magnetizing current keeps to flow via Figure 3. SEPIC during CCM operation when S1 is (a) on, and (b) off
the demagnetizing winding and D3, at this same time, D2 is
 The voltage value of L1 and L2 during the switch S1 on-time
can be detailed as: (a)
(2)

(3)
The voltage across L1 and L2 during the switch S1 off-time
can be detailed as:
(b)
(4)

(5)
According to the principle of volt-second balance, the volt-
second formula of L1 and L2 are shown as:
(6) Figure 4. Operating mode control (a) current (b) voltage

Since the supply system operates at peak current mode, the

(7) input supply provides a constant current into a large scale of
load voltage even at short circuit situation.
Where, DS is the duty cycle of S1.
D. Battery Model
(8)
The lithium-ion battery is recognized as an effective
technology among all the storage systems, it has highly
From Eq. (6-7) the ratio between VBAT and VIN can be performant characteristics i.e., running under wider operating
determined as: temperature range, it accepts higher injected current in short
time and a less self-discharge with the ability to be charged
(9) faster [23]. However, Fig. 5 illustrates the multiple operating
phases of charging a battery.
As per specifications, the inductor ripples current (∆IL) is a
critical factor to design an effective SEPIC. A rule of thumb is
to adopt 20 to 40% of the input current (IIN), as calculated in
Eq. (10). Moreover, too much ripples might cause an increased
rate of electromagnetic interference while too little would make
the PWM operation unstable [20].

(10)

For a more precise estimate of the input current, IIN is

divided by the estimated worst efficiency (η) when VIN
=VIN(min) and IOUT = IOUT(max). Furthermore, to secure the dc to
dc conversion using the SEPIC, it will be controlled by PI
controller in order to achieve fast and stable reaction to charge
the battery within safety measurements [21].
Figure 5. Battery charging phases
C. Voltage and peak current mode control
In accordance with the proposed application, the voltage The battery charger is frequently operated under three
and peak CMC are two regulating modes operating both to phases: first one is controlled with CMC where the reference
provide continuous control of the supply system, with only one current is set to Ibat-max to avoid overheating and overstrain
control at a time. They are designed to react fast, so accurate in phenomena [24].
order to fulfill the required charging rate [22]. The selected
type of control is started with programming the voltage mode Once Vbat achieves Vbat-max, means the battery charging
and peak current mode set points, which are available to the mode is switched to the second phase and the approach is
user. Besides, the maximum output voltage and the supply adjusted to the VMC where the output voltage is configured to
current can be controlled at any operating mode. Fig. 4 shows Vbat-max. At this case, the rate of Ibat is decreasing until it falls
both, the voltage and peak CMC. under Ibat-min where the third phase is attained, the strategy is
kept working at the same operation mode, except that the The applied control is chosen via various criteria, e.g., the
reference voltage is now adjusted to Vbat-float, this value is able real time monitoring of the charging process in order to get a
to avoid the worst deep self-discharge of the battery by reliable data for the configured mode control and for a display
generating a small charging current. purpose [25]. Table 1 depicts the setting data of each design.
TABLE I. 50W BATTERY CHARGER SETTINGS
III. SIMULATION RESULTS

To test the validity of these topologies and their strategies

of control, a design of 50W battery charger is modeled in
MATLAB/Simulink in order to set a comparative study
between the SEPIC and the forward converter. Furthermore,
the load is presented as a 12V/2Ah lithium-ion battery. Fig. 6
shows the two kinds of energy conversion schemes modeled in
Simulink environment. the control system is used for both
converters where the generated duty cycle based on PI control
with dynamic inputs of voltage/current and different strategies.

(a)

(b)

Figure 6. Designed converters in Simulink a) Forward b) SEPIC

 The simulation features are based, on the one hand, on a The program starts the process by sensing all the
comparative study between two topologies of dc/dc dynamic data i.e., voltage/current of the battery and the input
converter, on the other hand, on a mixed mode control and its current. The goal is to charge the lithium-ion battery within
effectiveness on the charging operation of lithium-ion its allowed limits in terms of state of charge, injected current
battery. However, the outcome is to gain as much as possible and the output charging voltage. Fig 9 shows the results of an
from each mode control where the battery electrical improved charging process using the previous algorithm.
characteristics will be preserved from eventual damages,
besides, the overheating phenomena is often caused by a
charging/discharging overcurrent. Following the project
specifications, the system is started with peak CMC to avoid
uncontrolled charging rate, after that the power management
block will switch the control to the VMC in order to achieve
more revenues from a large scale of power offered by the
50W charger. Fig. 7 illustrates the simulation results of a
battery charger based forward converter operated under peak
current mode control.

Figure 9. Waveforms of the hybrid control applied on forward converter

versus time

From t=0-4.6 min, the charging system is operated under

CMC where the charging current rate is set to IIN-max, after
t=4.6 min, the battery voltage reached VBat-max. In order to
stabilize the charging voltage rate, the control algorithm is
switched to the VMC, where the reference voltage is set to
VBat-max. At t=7.1min, the injected current goes down below
IIN-min, the VMC adjusted its reference voltage to VBat-float.
Figure 7. Waveforms of the CMC applied on forward converter versus time
The second part of the simulation section is to test this
As per specifications, a constant current control is algorithm on another topology of converter as the case o
provided by which the lithium-ion battery can be charged by SEPIC. Figure 10 depicts the results of a SEPIC running
a maximum power. As it can be seen, the charging battery under both the constant and the voltage mode control. The
voltage crosses its maximum limit at t=4.6 min, this state of charge of the battery identifies two rates of slope due
overvoltage will be increased in an uncontrolled way which to the changes in the charging power.
could causes harmful sides effects to the battery [26].
To take control of the selected charging operation, the
power management block is added to the platform in order to
implement the two modes of control following the guidelines
of an algorithm [27]. Fig 8 depicts the control approach
designed in Simulink using Matlab function block.

Figure 10. Waveforms of the hybrid control applied on SEPIC versus time
Figure 8. Control management algorithm of the system
 The above waveforms validate the theoretical approaches semiconductors (SiC and GaN) technology for power devices,”
of charging a battery using various modes of control Microelectronic Engineering, vol. 187–188, pp. 66–77, Feb. 2018.
performed at the same process. As results, all fixed goals are [12] M. Khalilian and E. Adib, “Soft-single-switched dual forward-flyback
PWM DC-DC converter with non-dissipative LC circuit,” 2015 23rd
achieved as preserving the battery from an eventual Iranian Conference on Electrical Engineering, May 2015.
overheating caused by overvoltage and overcurrent, and also [13] A. Hassoune, M. Khafallah, A. Mesbahi, and T. Bouragba, “Smart
the profitability in terms of a flexible control. topology of EVs in a PV-grid system based charging station,” 2017
International Conference on Electrical and Information Technologies
(ICEIT), Nov. 2017.
IV. CONCLUSION
[14] J. P. Torreglosa, P. García-Triviño, L. M. Fernández-Ramirez, and F.
In this paper, a control algorithm is presented in detail to Jurado, “Decentralized energy management strategy based on
charge a lithium-ion battery with quite a high level of predictive controllers for a medium voltage direct current
accuracy and stability. An effective real-time monitoring of photovoltaic electric vehicle charging station,” Energy Conversion
and Management, vol. 108, pp. 1–13, Jan. 2016.
the charging process leads to a reliable and a flexible modes
[15] C.-T. Tsai, T.-C. Liang, Y.-C. Kuo, and Y.-C. Luo, “An improved
of control via CMC and VMC. From the proposed schemes forward converter with PFC and ZVS features for split-phase charger
of converters, the control strategies achieved productive applications,” Computers & Electrical Engineering, vol. 51, pp. 291–
results while considering various criteria, such as the 303, Apr. 2016.
capacity of the battery, the rated power of the adopted [16] T.-C. Huang, Y.-G. Leu, Y.-C. Chang, S.-Y. Hou, and C.-C. Li, “An
charger and how the management algorithm would react energy harvester using self-powered feed forward converter charging
rapidly to the frequent changes in charging voltage/current. approach,” Energy, vol. 55, pp. 769–777, Jun. 2013.
The simulation results performed in Matlab/Simulink had [17] J. López, S. I. Seleme, P. F. Donoso, L. M. F. Morais, P. C. Cortizo,
validated the approach tested on two different converters i.e., and M. A. Severo, “Digital control strategy for a buck converter
operating as a battery charger for stand-alone photovoltaic systems,”
forward converter and SEPIC. Solar Energy, vol. 140, pp. 171–187, Dec. 2016.
[18] S. Venkatanarayanan and M. R. M. Nanthini, “Design and
REFERENCES Implementation of SEPIC and Boost Converters for Wind and Fuel
cell Applications,” International Journal of Innovative Research in
[1] A. Moro and L. Lonza, “Electricity carbon intensity in European Science, Engineering and Technology, vol. 3, no. 3, pp. 378-383,
Member States: Impacts on GHG emissions of electric vehicles,” Mar. 2014,
Transportation Research Part D: Transport and Environment, vol. 64,
pp. 5–14, Oct. 2018. [19] J. C. Rosas-Caro, V. M. Sanchez, R. F. Vazquez-Bautista, L. J.
Morales-Mendoza, J. C. Mayo-Maldonado, P. M. Garcia-Vite, and R.
[2] L. Benaaouinate, M. Khafallah, A. Mesbahi, A. Martinez, and T. Barbosa, “A novel DC-DC multilevel SEPIC converter for PEMFC
Bouragba, “Emulation of a Wind Turbine Using Induction Motor systems,” International Journal of Hydrogen Energy, vol. 41, no. 48,
Driven by Field Oriented Control,” International Journal of Energy pp. 23401–23408, Dec. 2016.
and Power Engineering, vol. 12, no. 3, 2018.
[20] K. Uddin, A. D. Moore, A. Barai, and J. Marco, “The effects of high
[3] K. Engeland, M. Borga, J.-D. Creutin, B. François, M.-H. Ramos, and frequency current ripple on electric vehicle battery performance,”
J.-P. Vidal, “Space-time variability of climate variables and Applied Energy, vol. 178, pp. 142–154, Sep. 2016.
intermittent renewable electricity production – A review,” Renewable
and Sustainable Energy Reviews, vol. 79, pp. 600–617, Nov. 2017. [21] Y. Guo, Z. Zhao, and L. Huang, “SoC Estimation of Lithium Battery
Based on AEKF Algorithm,” Energy Procedia, vol. 105, pp. 4146–
[4] A. Luo, Q. Xu, F. Ma, and Y. Chen, “Overview of power quality 4152, May 2017.
analysis and control technology for the smart grid,” Journal of
Modern Power Systems and Clean Energy, vol. 4, pp. 1–9, Jan. 2016. [22] D. A. Sbordone, B. Di Pietra, and E. Bocci, “Energy Analysis of a
Real Grid Connected Lithium Battery Energy Storage System,”
[5] N. Tewari and V. T. Sreedevi, “A novel single switch dc-dc converter Energy Procedia, vol. 75, pp. 1881–1887, Aug. 2015.
with high voltage gain capability for solar PV based power generation
systems,” Solar Energy, vol. 171, pp. 466–477, Sep. 2018. [23] G. Zubi, R. Dufo-López, M. Carvalho, and G. Pasaoglu, “The
lithium-ion battery: State of the art and future perspectives,”
[6] A. Hassoune, M. Khafallah, A. Mesbahi, and D. Breuil, “Electrical Renewable and Sustainable Energy Reviews, vol. 89, pp. 292–308,
design of a photovoltaic-grid system for electric vehicles charging Jun. 2018.
station,” 2017 14th International Multi-Conference on Systems,
Signals & Devices (SSD), Mar. 2017. [24] V. Lystianingrum, B. Hredzak, and V. G. Agelidis, “Multiple model
estimator based detection of abnormal cell overheating in a Li-ion
[7] S. Chalise, J. Sternhagen, T. M. Hansen, and R. Tonkoski, “Energy battery string with minimum number of temperature sensors,” Journal
management of remote microgrids considering battery lifetime,” The of Power Sources, vol. 273, pp. 1171–1181, Jan. 2015.
Electricity Journal, vol. 29, no. 6, pp. 1–10, Jul. 2016.
[25] A. Hassoune, M. Khafallah, A. Mesbahi, and T. Bouragba “Power
[8] S. Rivera and B. Wu, “Electric Vehicle Charging Station With an Management Strategies of Electric Vehicle Charging Station Based
Energy Storage Stage for Split-DC Bus Voltage Balancing,” IEEE Grid Tied PV-Battery System,” International Journal of Renewable
Transactions on Power Electronics, vol. 32, no. 3, pp. 2376–2386, Energy Research, vol. 8, no. 2, pp. 851-860, Jun. 2018.
Mar. 2017.
[26] S. Guo, R. Xiong, K. Wang, and F. Sun, “A novel echelon internal
[9] “ABB Launching 350 kW EV Fast Charger At Hannover Messe heating strategy of cold batteries for all-climate electric vehicles
[Online]. Available: https://cleantechnica.com/2018/04/25/abb- application,” Applied Energy, vol. 219, pp. 256–263, Jun. 2018.
launching-350-kw-ev-fast-charger-at-hannover-messe/ [Accessed: 25-
April-2018]. [27] A. Hassoune, M. Khafallah, A. Mesbahi, L. Benaaouinate, and T.
Bouragba, “Control Strategies of a Smart Topology of EVs Charging
[10] L. Pan and C. Zhang, “Performance Enhancement of Battery Charger Station Based Grid Tied RES-Battery,” International Review of
for Electric Vehicles Using Resonant Controllers,” Energy Procedia, Electrical Engineering (IREE), vol. 13, no. 5, p. 385-396, Oct. 2018.
vol. 105, pp. 3990–3996, May 2017.
[11] F. Roccaforte, P. Fiorenza, G. Greco, R. Lo Nigro, F. Giannazzo, F.
Iucolano, and M. Saggio, “Emerging trends in wide band gap