Overview of multi-stage charging strategies for Li-ion batteries

Abstract

To reduce the carbon footprint in the transportation sector and improve overall vehicle efficiency, a large number of electric vehicles are being manufactured. This is due to the fact that environmental concerns and the depletion of fossil fuels have become significant global problems. Lithium-ion batteries (LIBs) have been distinguished themselves from alternative energy storage technologies for electric vehicles (EVs) due to superior qualities like high energy and power density, extended cycle life, and low maintenance cost to a competitive price. However, there are still certain challenges to be solved, like EV fast charging, longer lifetime, and reduced weight. For fast charging, the multi-stage constant current (MSCC) charging technique is an emerging solution to improve charging efficiency, reduce temperature rise during charging, increase charging/discharging capacities, shorten charging time, and extend
the cycle life. However, there are large variations in the implementation of the number of stages, stage transition criterion, and Crate selection for each stage. This paper provides a review of these problems by compiling information from the literature. An
overview of the impact of different design parameters (number of stages, stage transition, and C-rate) that the MSCC charging
techniques have had on the LIB performance and cycle life is described in detail and analyzed. The impact of design parameters on lifetime, charging efficiency, charging and discharging capacity, charging speed, and rising temperature during charging is presented, and this review provides guidelines for designing advanced fast charging strategies and determining future research gaps.

Full text

Review
Overview of multi-stage charging strategies for Li-ion batteries
Muhammad Usman Tahir
⇑
, Ariya Sangwongwanich, Daniel-Ioan Stroe, Frede Blaabjerg
AAU Energy, Aalborg University, Aalborg DK-9220, Denmark
article info
Article history:
Received 4 April 2023
Revised 21 May 2023
Accepted 21 May 2023
Available online 5 June 2023
Keywords:
Multi-stage constant current (MSCC)
charging
Electric vehicles (EVs)
Li-ion batteries (LIBs)
Fast charging strategies
abstract
To reduce the carbon footprint in the transportation sector and improve overall vehicle efficiency, a large
number of electric vehicles are being manufactured. This is due to the fact that environmental concerns
and the depletion of fossil fuels have become significant global problems. Lithium-ion batteries (LIBs)
have been distinguished themselves from alternative energy storage technologies for electric vehicles
(EVs) due to superior qualities like high energy and power density, extended cycle life, and low mainte-
nance cost to a competitive price. However, there are still certain challenges to be solved, like EV fast
charging, longer lifetime, and reduced weight. For fast charging, the multi-stage constant current
(MSCC) charging technique is an emerging solution to improve charging efficiency, reduce temperature
rise during charging, increase charging/discharging capacities, shorten charging time, and extend the
cycle life. However, there are large variations in the implementation of the number of stages, stage tran-
sition criterion, and C-rate selection for each stage. This paper provides a review of these problems by
compiling information from the literature. An overview of the impact of different design parameters
(number of stages, stage transition, and C-rate) that the MSCC charging techniques have had on the
LIB performance and cycle life is described in detail and analyzed. The impact of design parameters on
lifetime, charging efficiency, charging and discharging capacity, charging speed, and rising temperature
during charging is presented, and this review provides guidelines for designing advanced fast charging
strategies and determining future research gaps.
Ó2023 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published
by ELSEVIER B.V. and Science Press. This is an open access article under the CC BY license (http://creati-
vecommons.org/licenses/by/4.0/).
Muhammad Usman Tahir received a bachelor’s degree
in electrical engineering from Government College
University (GCU) Lahore and M.Sc in electrical engi-
neering from Lahore University of Management Sciences
(LUMS) in 2017 and 2019, respectively. From June 2019
to January 2022, he worked as a research assistant and
teaching assistant at LUMS. He is pursuing a Ph.D. in the
Department of Energy at Aalborg University. His
research interests include Li-ion battery charging tech-
niques, power electronics converters, reliability engi-
neering, and renewable energy systems.
Ariya Sangwongwanich received the M.Sc. and Ph.D.
degrees in energy engineering from Aalborg University,
Aalborg, Denmark, in 2015 and 2018, respectively. He
was a Visiting Researcher with RWTH Aachen, Aachen,
Germany from September 2017 to December 2017. He is
currently a assistant professor with the Department of
Energy, Aalborg University. His research interests
include control of grid-connected power converters,
photovoltaic systems, reliability in power electronics,
and multilevel converters. Dr. Sangwongwanich was a
recipient of the Danish Academy of Natural Sciences’ Ph.
D. Prize and the Spar Nord Foundation Research Award
for his Ph.D. thesis in 2019.
https://doi.org/10.1016/j.jechem.2023.05.023
2095-4956/Ó2023 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by ELSEVIER B.V. and Science Press.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
⇑
Corresponding author.
E-mail address: mut@energy.aau.dk (M. Usman Tahir).
Journal of Energy Chemistry 84 (2023) 228–241
Contents lists available at ScienceDirect
Journal of Energy Chemistry
journal homepage: www.elsevier.com/locate/jechem

Daniel-Ioan Stroe received the Dipl.-Ing. degree in
automatics from Transilvania University of Brasov,
Romania, in 2008, and M.Sc. degree in wind power
systems from Aalborg University (AAU), Aalborg, Den-
mark, in 2010. He has been with Aalborg University
since 2010, from where he obtained his Ph.D. degree in
lifetime modelling of Lithium-ion batteries in 2014.
Currently, he is an Associate Professor with Departe-
ment of Energy, Aalborg University, where he leads the
Batteries research group and the Battery Systems Test-
ing Lab. He was a Visiting Researcher with RWTH
Aachen, Germany, in 2013. His current research inter-
ests are in the area of energy storage systems for grid and e-mobility, Lithium-based
batteries testing, modelling, diagnostics and their lifetime estimation.
Frede Blaabjerg received the Ph.D. degree in electrical
engineering from Aalborg University, Aalborg, Denmark,
in 1995. He was with ABB-Scandia, Randers, Denmark,
from 1987 to 1988. He became an Assistant Professor in
1992, an Associate Professor in 1996, and a Full Profes-
sor of power electronics and drives in 1998 with the
Department of Energy, Aalborg University. In 2017, he
became a Villum Investigator. He is a Honoris Causa at
University Politehnica Timisoara (UPT), Romania, and
Tallinn Technical University (TTU), Estonia. He has
published more than 600 journal articles in power
electronics and its applications. He is the coauthor of
four monographs and an editor of ten books in power electronics and its applica-
tions. His current research interests include power electronics and its applications in
wind turbines, PV systems, reliability, harmonics, and adjustable speed drives. Dr.
Blaabjerg has received 32 IEEE Prize Paper Awards, the IEEE PELS Distinguished
Service Award in 2009, the EPE-PEMC Council Award in 2010, the IEEE William E.
Newell Power Electronics Award 2014, the Villum Kann Rasmussen Research Award
2014, the Global Energy Prize in 2019, and the 2020 IEEE Edison Medal. He was the
Editor-in-Chief of the IEEE TRANSACTIONS ON POWER ELECTRONICS from 2006 to
2012. He has been a Distinguished Lecturer of the IEEE Power Electronics Society
from 2005 to 2007 and of the IEEE Industry Applications Society from 2010 to 2011
as well as 2017 to 2018. From 2019 to 2020, he served as a President for IEEE Power
Electronics Society. He is the Vice-President of the Danish Academy of Technical
Sciences. He was nominated in 2014 and 2019 by Thomson Reuters to be among the
most 250 cited researchers in Engineering in the world.
1. Introduction
By the end of 2021, there were approximately 16.5 million elec-
tric vehicles (EVs) on the road worldwide, which was an enormous
growth compared to 2020 (10 million EVs were on roads). By the
first quarter of 2022, 2 million EVs had been sold, a 75% increase
over the same period in 2021 [1]. EVs have grown in popularity
in recent years due to their benefits in terms of energy efficiency
and solution to the current environmental challenges, such as
emissions, reducing reliance on fossil fuels, and air pollution in
the cities. More than 20 countries have announced electrification
goals or internal combustion engine (ICE) bans for vehicles. Eight
countries, including the European Union (EU), have committed to
net-zero emissions by 2050 [1,2]. The EVs must be mass-
produced and sold in massive amounts to achieve this goal. The
key components of EVs are power converters and lithium-ion bat-
teries (LIBs). The LIBs stand out significantly compared to other
energy storage technologies due to their high energy and power
density, high efficiency, low maintenance cost, no memory effect,
and low self-discharge rate [3–7]. From the user’s perspective,
the issues with EVs include their overall mileage, charging speed,
range anxiety, and safe and reliable operations during driving. All
these features depend strongly on the battery charging and dis-
charging capacity, lifetime, charging time, and charging efficiency
[8]. Traditionally, the current rate (C-rate) influences the
performance-degradation behavior of LIBs. Thus, the charging
method impacts the performance and lifetime parameters of the
LIB [9]. On the other hand, the battery discharging is determined
by the consumer’s energy consumption behavior. To improve the
lifetime and performance characteristics of LIBs, research on charg-
ing strategies is a crucial issue topic.
A suitable charging protocol is required for the optimal charging
of LIBs. During the charging of LIBs, the battery charger controls the
voltage, current, and/or power of LIBs [10]. Fast charging tech-
niques for EV applications generally aim to achieve the optimal
balance between the two contradictory objectives of reducing
charging time and extending the lifetime [11,12]. However, life-
time must be carefully taken into account because high charge cur-
rent rates are usually one of the main factors that cause
degradation and low performance of LIBs [11–14]. The cause is
attributed to several aging mechanisms, including solid electrolyte
(SEI) growth [15–17], lithium plating [16,18–21], and mechanical
degradation [17,22]. These aging mechanisms are highly depen-
dent on different factors, including C-rate, temperature, state of
charge (SOC), and state of health (SOH). The charging parameters
vary depending on the operating conditions. The theoretical illus-
tration of the charging process is shown in Fig. 1 [23,24]. During
the charging process, electrons are extracted from the cathode
and moved toward the anode through the external circuit by the
charger. Meanwhile, Li
+
ions are deintercalated from the cathode
electrode and moved to the anode electrode by passing through
the electrolyte [25]. The charging process has three steps: (1) the
diffuse-out of the Li
+
ions from the cathode, (2) the diffusion of sol-
vated Li
+
ions into the electrolyte, and (3) the de-solvation of Li
+
ions by passing through the SEI and intercalation into the
Fig. 1. The charging procedure for a standard Li-ion battery [24]. During the
charging process, the solvated lithium ions (Li
+
) are intercalated to the negative
electrode (anode) after being de-intercalated from the positive electrode (cathode).
The Li
+
is turning into Li after accepting the electrons from the electrode.
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
229

interlayer of graphite [18,26]. These are the ideal LIB working con-
ditions. Nonetheless, LIBs are subjected to severe working condi-
tions in real-world applications, which have a substantial impact
on the side reactions during ion transport [19]. Therefore, the LIB
performance and lifetime are substantially impacted by charging
conditions [26]. Furthermore, the thermal behavior of the LIB is
significantly reliant on the charging/discharging current. Therefore,
the considerable heat generation raises safety concerns. Also, opti-
mizing the charging strategy several times during battery opera-
tions can also extend the remaining useful battery lifetime [27].
Therefore, a suitable charging protocol is required for the optimal
charging of LIB, which will positively impact the performance
and lifetime of LIBs.
The constant current constant voltage (CCCV) is the standard
charging method for LIBs [28–32]. This charging method consists
of two stages, as depicted in Fig. 2 (a). In the first stage, the LIB
is charged with constant current (CC), I
ch
, until the maximum volt-
age V
max
is reached. In the second stage, the LIB is charged with
constant voltage (CV) until either the I
ch
reaches the predefined
cut-off current value I
end
(typically, it is 5% of the nominal current)
or a predefined maximum time t
max
is reached. However, I
ch
has
the greatest impact on the charging rate, whereas V
max
and I
end
define the capacity of charged battery [11]. The manufacturer usu-
ally defines the I
ch
rating, which is dependent on the chemistry of
the LIB. Furthermore, higher C-rate charging has a negative effect
on LIB health [33]. On the other hand, generally, CV mode takes a
lot of the charging time and only has a limited contribution to
the amount of charged capacity [34]. CV stage is very dependent
on the C-rate employed, temperature, and the degradation level
of the battery. Also, the diffusion limiting C-rate is an additional
restriction that limits the diffusion process within the electrodes
at higher C-rates, affecting the capacity and performance of LIB
[35]. Fast Li
+
ion diffusivity in the active materials is recognized
as one of the significant factors needed for fast charging [36]. Addi-
tionally, charging at high C-rates will lead to lithium plating in bat-
teries, especially at low temperatures or at high SOC [19]. From the
electrochemistry perspective, Li
+
builds up on the surface of the
anode and transforms into lithium metal when the rate of diffusion
of Li
+
ions to the anode material surface is faster than the rate of
diffusion inside the material [37,38]. These conditions limit the
charge transfer kinetics in the electrolyte and solid-state diffusion,
causing the anode potential to drop below the lithium metal
potential. This phenomenon, known as lithium plating, can nega-
tively impact the LIB performance and cycle life. Therefore, various
authors have proposed different charging techniques that restrain
the lithium plating on the anode surface, prolonging the battery
life, reducing charging time, decreasing temperature rise, and
improving the charging efficiency [18,29,38–40].
Boost charging (BC) is one technique to improve the charging
speed of the LIB compared to the CCCV method [11]. BC is a variant
of CCCV charging that includes a higher CC or constant power (CP)
period at the start of the charging period [41]. Because the LIBs are
less sensitive to lithium plating at low SOC, this additional boost
interval will minimize the charging time without compromising
the cycle life. The BC charging strategy is shown in Fig. 2(b). The
figure illustrates that the cell is first charged with a high current
I
boost
until enough charge is supplied into the cell (e.g., 40% SOC).
The conventional CCCV procedure is then applied after the I
boost
interval. The charging speed can be adjusted with the charging cur-
rent level I
boost
. Furthermore, the same conditions are applied to
alter capacity utilization and charging rate as CCCV charging.
When the boost interval is between 20%–60% SOC, the author in
ref. [11] found that boost charging does not affect LIB capacity fade,
thereby the aging, while minimizing the charging time [42,43].
This charging approach could be effective in EV applications for
fast charging [28].
The pulse charging technique has been proposed by various
authors to reduce the charging time, decrease temperature rise,
and prolong the LIB lifespan [34,44,45]. Different types of current
pulses exist based on the current frequency, amplitude, and duty
cycle [4,44]. Several pulse charging patterns were proposed in
the literature as a substitute for CCCV charging. The pulse charging
is based on the periodic variation of different current pulses with
variable current rates and directions. The charging current can be
suspended, increased, decreased, or replaced by short discharging
pulses for a specific period. As illustrated in Fig. 2(c), the most typ-
ical pulse charging approach is the positive pulse current (PPC)
with some relaxation periods, where the design parameters are
the peak charging current (I
p
), the on-time of the pulse (t
p
), the
relaxation interval time (t
r
), and the total period of the pulse (T),
respectively. According to ref. [36], the PPC can extend the LIB
lifetime by 60% at low frequency (0.05 Hz) and by 105% at high
Fig. 2. Different charging methods for lithium-ion battery. (a) Standard CCCV method; (b) BC method; (c) PPC method; (d) NPC method followed by PCC; (e) MSCC charging
strategy.
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
230

frequency (2 kHz) [46]. The Taguchi orthogonal arrays technique
was used in ref. [47] to find the best pulse charging parameters
that improve LIB charge and energy efficiency while reducing
charging time. It was discovered that operating a PPC with ideal
parameters reduced charging time by 47.6% and enhanced LIB
charge and energy efficiency by 1.5% and 11.3%, respectively. The
author in ref. [48] presented that the PPC reduces the charging
time of 100 mA h and 45 mA h LIB by 37.35% and 15.56%, respec-
tively. Another pulse charging strategy is negative pulse current
(NPC), as shown in Fig. 2(d). In this case, the parameters I
p
,I
n
,t
p
,
t
r
,t
n
, and Tare abbreviations for the peak charging current, peak
discharging current during charging, the on-time of the pulse, the
relaxation interval time, negative pulse time, and the total duration
of the pulse, respectively. According to the authors in ref. [49], the
NPC can increase active material utilization, resulting in higher dis-
charge capacity and a longer lifetime. According to the authors in
ref. [10], NPC with lower frequency and amplitude improves charg-
ing capacity by about 3.5% compared to NPC with higher amplitude
and frequency. Therefore, it was established that pulse charging
with a variety of modifications is slightly beneficial for LIB life
cycles, charging time reduction, and increased discharge capacity
[10,34,44–49].
Multi-stage constant current (MSCC) charging is another charg-
ing strategy that has been proposed by various researchers to
reduce the charging time, enhance the charging efficiency, and pro-
long the LIB lifetime [30,50–57].Fig. 2(e) depicts the MSCC charg-
ing for LIBs. In the literature, various authors proposed a variable
number of stages to charge a LIB. In refs. [30,50], the author con-
ducted various experiments and discovered that the three-stage
charging technique is more superior to the standard method
(CCCV) in terms of reducing charging time and enhancing charging
and discharging capacity. The authors of ref. [58] conducted exper-
iments to determine the best C-rate for four-stage charging. Com-
pared to the standard charging technique (CCCV), the authors
determined that the optimal four-stage charging reduces charging
time, extends LIB lifetime, and increases charge capacity without
inducing temperature rise. In ref. [52], the author found the cur-
rents of the five-stage charging using the grey wolf optimization
method, compared the results with the CCCV method, and discov-
ered that utilizing MSCC charging reduced charging time by 5.33%,
improved charging efficiency by 12.54%, and extended the life
cycle by 79.6%. The authors of ref. [59] used a five-stage MSCC
charging approach and identified the criteria to avoid lithium plat-
ing. The results indicate that the optimal MSCC has a lower effect
on lithium plating and enables faster charging speeds as well as
better capacity retention compared to the CCCV charging method.
In ref. [60], the effects of MSCC charging at three different ambient
temperatures were investigated. Compared to CCCV, MSCC
decreases the charging time and increases the charging capacity
at low and high temperatures. Many authors have proposed differ-
ent numbers of charging stages to evaluate the charging time,
charging efficiency, life cycles, and temperature rise. However,
there is no consensus on the effect of MSCC on LIBs because studies
have used different test goals, research objectives, and testing pro-
cedures. Therefore, it is necessary to conduct a comprehensive lit-
erature study, including comparison and drawing conclusions
regarding the effect of MSCC strategies on LIBs, which is essential
to be able to lead research into the optimal charging strategy.
Each charging technique has advantages and disadvantages of
its own. Although CCCV charging is relatively simple to implement,
it is unsuitable for fast charging. Similarly, PPC and NPC charging
methods decrease charging time but are expensive and compli-
cated. However, the MSCC charging approach is simple to use.
The limitation of MSCC is the need to predict the SOC or voltage
level for their utilization precisely. Several review publications
and their contribution to the evaluation of the charging approach
are presented in Table 1. The publications have focused on pulse
charging, optimal charging strategies, and the dynamics of materi-
als during fast charging. Therefore, this study aims to conduct a
thorough analysis of the MSCC charging approach and its influence
on the lifetime and performance characteristics of LIBs. This paper
provides a comprehensive overview of the MSCC charging strategy.
The existing MSCC charging modes are discussed in Section 2. The
technique of transitioning into different stages is discussed in
Section 3. The effects of MSCC charging on the charging time,
charging efficiency, charging/discharging capacity, temperature
rise, safety, and LIB lifetime are discussed in Section 4.Section 5
presents the analysis and discussion based on the findings. Sec-
tion 6 provides the conclusions for the analysis done.
2. Multi-stage constant current charging for LIBs
Fast charging is one of the key solutions for overcoming the
range anxiety issue that has limited the deployment of EVs
[65,66]. The development of a charging strategy is critical for
improving the EV user experience as well as increasing the charg-
ing efficiency, reducing charging time, extending the lifetime, and
improving the safety of LIBs [67–69]. One of the primary charging
strategies that can address the previously mentioned aspects is
MSCC charging. According to the design requirements of battery
charging, the CC during charging is divided into several currents
with decreasing C-rates in the MSCC charging strategy. The
decreasing CC rate considerably minimizes lithium precipitation
from the negative electrode of the battery during continuous
charging with a large current, as shown in Fig. 2(e) [55]. Moreover,
as compared to the CCCV strategy, the MSCC charging approach
eliminates the constant voltage charging stage. Thus, the elec-
trolyte oxidation process produced by the high voltage is reduced
[70]. As a result, the MSCC charging approach has attracted the
large interest of many researchers.
Table 1
Summary of several research articles and their contribution to the evaluation of fast
charging.
Ref. Strategies Contribution
[4] Pulse charging Comprehensive analysis of pulse
charging and its effect on the
performance and lifespan of LIBs.
[28] Fast charging
strategies
Review of fast-charging techniques and
control limitations.
[31] CCCV Examining the influence of fast charging
on the graphite anode as a result of Li-
plating and the structural instability of
layered lithium-metal oxide at the
cathode.
[61] Pulse charging A summary and analysis of the impact of
pulse charging on LIBs performance.
[39] Charging
optimization
Thorough overview of charge
optimization strategies, including a
general framework and a control
protocol.
[62] Charging and
discharging
strategies
Covering the systemic charging and
discharging patterns of EVs at the level
of the grid and providing
recommendations.
[63] Extreme fast
charging
Overview of the fast-charging strategies
and discussion on the limitations and
challenges based on material
perspectives.
[64] Optimal charging
strategy
Review and explanation of the charging
approach based on the dynamics and
control of the materials in the batteries.
This work MSCC A comprehensive review on the MSCC
charging strategies and their impact on
the performance and lifetime of LIBs.
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
231

Based on the multi-stage concept, a varietyofchargingstages
ranging from three to ten were proposed in the literature, and the
results are quite diverse. In addition, the criteria for transitioning
from one stage to another varies, and the current rating varies from
stage to stage based on the method used to determine the C-rate.
There are four parameters that are usually used to analyze the perfor-
manceofLIBs:(1) charging time, (2) charging efficiency, (3) charge/
discharging capacity, and (4) temperature rise during charging.
2.1. Charging time
The charging time is determined by the charging C-rate; the
higher the C-rate, the shorter the charging time. However, increas-
ing the C-rate during the charging process has a detrimental effect
on charging efficiency and cycle life. Higher charging current,
within the manufacturer’s specified limit, in the low SOC level
has a less detrimental impact on LIB life cycles and performance
parameters [71]. When the SOC exceeds a particular threshold
(usually it is 40%–50% of SOC), a higher charging current has a neg-
ative impact owing to polarization. Polarization occurs when Li
+
is
intercalated on the anode, and as a result, the capacity fade occurs.
In the MSCC charging strategy, the charging current reduces from a
higher C-rate to a lower C-rate to reduce the charging time and cre-
ate a less polarization effect on LIBs [71].
2.2. Charging efficiency
Charging efficiency is one of the key performance indicators for
the LIB charging procedure. Charging efficiency is defined as the
ratio of discharging capacity extracted from the charged LIB to
the charging capacity of LIB during charging. The charging effi-
ciency is computed using the following Eq. (1).
g
ð%Þ¼ I
d
t
d
P
n
i¼1
I
c
t
c
100 ð1Þ
In Eq. (1),n,I
d
,t
d
,I
c
, and t
c
show the number of stages, discharg-
ing current, discharging time, charging current, and charging time,
respectively. In the CCCV method, the capacity loss during the
charging procedure increases as the C-rate increases. The same
concept is applied to the MSCC charging strategy for higher C-
rates. However, the energy loss is reduced due to the varied num-
ber of charging stages in comparison to the CCCV method. It is
essential to reduce energy loss to improve charging efficiency. In
some cases, the charging time and energy loss conflict with each
other [72,73]. In ref. [74], it is shown that the charging time is
reduced by 34% compared to CCCV using the SOC-based four-
stage charging strategy, but the energy efficiency is reduced by
0.6% as well. In another case, the MSCC charging strategy with volt-
age cut-off transition criteria has a positive impact on energy effi-
ciency with optimal charge pattern [51,75,76].
2.3. Charge/discharge capacity
Another performance parameter is the charge/discharge capac-
ity. The entire amount of charge that is stored in the LIB during
charging is known as the charging capacity. Due to operational
conditions such as ambient temperature, last-stage charging cur-
rent, and end-of-charge criteria, the charge capacity will vary.
The LIBs are not fully charged at 0 °C or negative temperature
[60]. Similarly, the discharging capacity is the total amount of
energy that is drawn during discharging of LIBs. Thus, the normal-
ized charge and discharge capacity can be evaluated using Eqs. (2)
and (3) [51].
NCC ¼C
cMSCC
C
ccccv
ð2Þ
NDC ¼C
dMSCC
C
dcccv
ð3Þ
Here NCC, NDC, C
c-MSCC
,C
c-cccv
,C
d-MSCC
, and C
d-cccv
are the abbre-
viations of normalized charge capacity, normalized discharge
capacity, charge capacity during MSCC charging, charge capacity
during CCCV charging, the discharge capacity of charged LIB
(charged with MSCC charging strategy), and discharge capacity of
charged LIB (charged with CCCV method), respectively. The charge
and discharge capacity are compared with those of the standard
charging method (CCCV). MSCC charging strategy has a positive
impact on the charge and discharge capacity of LIB compared to
the CCCV method [51,55,60].
2.4. Temperature rise
The rise in temperature during charging has a negative impact
on the performance and lifetime of LIB. During fast charging, signif-
icant heat is generated [40,77]. The heat produced by ohmic loss is
proportional to the square of the root mean square (RMS) current.
As the current increases, the temperature of LIB rises. In addition,
LIBs are occasionally subject to thermal runaway due to their
tremendous heat generation [78,79]. Therefore, the temperature
rise during charging has a negative impact on the overall perfor-
mance and lifetime of LIBs.
2.5. Cycle life
Numerous techniques can be used to determine the cycle life of
LIB. Generally, the aging test is conducted at various charging pro-
files of the MSCC charging strategy and compared with the CCCV or
CC method. The cycle life depends upon the chemistry and con-
stituents (electrolyte, electrodes, and separator materials) of the
LIB. Depending on the operating conditions (temperature, current,
and voltage), the active components of LIB degrade with time, and
cycle life degrades.
Various works in the literature focus on particular performance
parameters (charging time, charging efficiency, charge/discharge
rate, and temperature rise) instead of all of them. Also, the findings
of each article are different from the others based on the chemistry
of the cell and the method used to find the C-rate, and the number
of multiple charging stages.
Different methods are used to assess the lifetime and perfor-
mance. The optimal C-rate is found using a variety of techniques,
including empirical approaches, analytical techniques, and opti-
mization methods. Various methods of transitioning from one
charging stage to the next stage are also proposed and validated.
The conceptual diagram for the MSCC charging strategy is shown
in Fig. 3, along with the methods and strategies used to do a tran-
sition from one stage to the next one and the parameters used to
evaluate the performance. From the literature, the MSCC charging
strategy improves performance metrics and lifetime. For example,
in ref. [50], the authors conducted numerous experiments and dis-
covered that the three-stage charging reduced the charging time
and increased the charging capacity when compared to the CCCV
method. The authors of ref. [58] concluded that using a four-
stage charging technique results in a shorter charging time, less
charge capacity loss, and longer cycle life for LIBs. The four-stage
charging strategy was also used to study the impact of weighting
parameters on the Taguchi method [30]. The equal weighting fac-
tor technique, according to the authors, increased the charging effi-
ciency and decreased charging time. The five-stage charging
increased the charging efficiency by 2.8% and decreased the tem-
perature rise by 9.3 °C compared to the CCCV method [76]. Using
a five-stage charging method, the authors of ref. [52] concluded
that a multi-stage charging strategy reduces charging time while
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
232

simultaneously increasing cycle life and charging efficiency. The
authors of ref. [80] used an eight-stage charging technique and
determined that multi-stage charging minimizes temperature rise
and charging time. The author in ref. [51] utilized the ten-stage
charging strategy, and the result demonstrated that the
multi-stage charging strategy increases charging capacity at lower
temperature while reducing charging capacity compared to the
CCCV method. Various authors suggested varying numbers of
stages for charging the LIBs. Additionally, different authors used
different criteria to transition between stages, and different C-
rates are proposed. These topics are covered in further details in
Section 3.
3. Design of charging pattern
The optimal design of the charging pattern is critical. To identify
the optimal charging pattern, several authors have employed a vari-
ety of techniques. Researchers employed different charging tech-
niques to find the C-rates for the MSCC charging, i.e., analytical
methods, optimization techniques, and empirical approaches. Below
is a thorough review of various techniques to find the C-rate for
charging the LIBs. In addition, the transition from one stage to the
next stage is another significant variation across the literature. Vari-
ous authors proposed different ways of transitioning from one stage
to another one. The stage transition criteria are discussed below.
Fig. 3. Conceptual design for the MSCC charging strategies, along with mentioned techniques for determining the C-rate for each stage and criteria for transitioning of stages
from one to the next one.
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
233

3.1. Stage transition criteria
The transition from one stage to the next one is determined by
meeting one of the following four criteria, found in the relevant
studies.
3.1.1. Time-based transition
Time is one parameter to switch from one stage to the next
stage, as illustrated in Fig. 4(a). Time-based transitions are simple
to implement, but there is no standard method for determining
when stages are shifted from one to the next one. The three-
stage charging strategy was chosen to charge the LIB up to 80%
of SOC in less than 40 min [71]. The time interval is chosen based
on the SOC intervals. The best-chosen group has time lengths of 10,
12, and 14 min for three stages with SOCs of 0%–30%, 30%–60%, and
60%–80%. The C-rates were selected randomly, and several experi-
ments were conducted to identify the optimal pattern (1.8 C, 1.5 C,
and 0.9 C). The findings revealed that MSCC charging improves
cycle life and significantly shortens the charging time compared
to the CC method. Also, the MSCC charging strategy has a lower
polarization and lithium plating effect compared to the CC method.
However, no standard procedure is proposed to find the optimal C-
rate for each stage.
3.1.2. SOC-based transition
To transition from one stage to another one, different studies
proposed distinct numbers of SOC stages. The accurate estimation
of SOC is a challenging task. The SOC-based transition was used to
shift from one stage to another one [50,74,80–83], as shown in
Fig. 4(b). The specific number of stages used in various papers is
summarized in Table 2. The number of stages was selected ran-
domly and applied to the LIB to determine the charging time,
charging efficiency, charge and discharge capacity, and tempera-
ture rise during charging. The results revealed that the proposed
MSCC strategy with SOC-based transition can reduce the charging
time [50,80–83], improve charging efficiency [82], enhance the
charge/discharge capacity [50], reduce the temperature rise [82],
and extend the LIB lifetime [83] in comparison to the standard
method (CCCV). In ref. [50], the authors performed experiments
on different SOC intervals. As higher the first-stage SOC interval
as much lower the charging time and the higher the temperature
rise. Furthermore, the authors of ref. [74] concluded that the
four-stage SOC-based MSCC charging approach reduced charging
time (roughly 15.3%), slightly decreased charging efficiency
(0.4%), and slightly reduced the temperature rise when compared
to the equivalent CCCV method. In addition, authors used post-
mortem analyses to investigate the impact of MSCC on graphite
exfoliation and crystallization damage [83]. The results demon-
strated that the MSCC reduced the SEI layer growth of anode, low-
ered internal resistance, and extended the LIB lifetime.
SOC estimation is challenging due to significant variations in
the battery parameters (voltage, current, and operating tempera-
ture during its lifespan owing to aging and nonlinear behavior).
In addition, the SOC-based stage transition is expensive due to its
computational cost and burden [84] and a little bit complex to
implement due to various parameters in the LIB operations. There
are various methods in the literature for calculating the SOC; how-
ever, each strategy has certain advantages and drawbacks [85].
Therefore, it can be challenging to apply the SOC-based transition
in practical application due to the need for relying on a parameter,
which is not accurately measurable and needs to be estimated.
3.1.3. Voltage threshold-based criteria
The voltage threshold-based transition is another criterion for
designing charging transitioning between two consecutive stages.
In this case, a different voltage threshold is selected to modify
the stage and charging current value, as shown in Fig. 4(c). The
authors employed the voltage threshold-based stage transition
[20,21,29]. The voltage thresholds ranged from 3.6 to 4.2 V since
the author selected to use gradually decreasing voltage increments
between each stage. The finding revealed that using a five- and
ten-stage voltage threshold-based technique reduced the charging
Fig. 4. Theoretical illustration of current (blue) and voltage (red) profiles verses time during MSCC charging strategies for transition criteria from one stage to the next stage.
(a) Time-based transition between stages; (b) SOC-based transition; (c) threshold Voltage based transition; (d) cut-off voltage-based transition.
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
234

time while the lifetime was unaffected [21,29]. Similarly, the
authors employed four- and five-stage MSCC charging strategies
to check its effect on the lithium plating [20]. The findings revealed
that high C-rate charging strategies have more lithium plating dur-
ing charging than low C-rate charging strategies. The author
selected voltage-threshold values randomly. No particular method
is used to determine the threshold voltage at each stage. Neverthe-
less, the threshold-based stage transition technique is easier to
implement than the SOC-based transition criteria.
3.1.4. Cut-off voltage-based criteria
Cut-off voltage-based criteria are the most utilized criterion in
the literature for transitioning from one stage to the next one. An
illustration of this strategy is shown in Fig. 4(d). The manufacturer
predetermines the cut-off voltage value. Usually, it is 4.2 V for
cobalt-based LIBs. In this method, once the cut-off voltage is
achieved while charging, the current is reduced (stage transition-
ing), and charging continues until the cut-off voltage is reached
again. This procedure is repeated until the predefined number of
stages is reached. Therefore, it is simple and easy to implement.
In literature, various authors implemented different stages, like
four stages [30,58,73,86], five stages [51,52,75,76,87–90], and ten
stages [60] to check the LIB lifetime and performance. The result
shows that the MSCC charging strategy can improve the charging
efficiency [30,51,52,76,88–90], reduce the charging time
[30,52,58,60,73,76,86–90], enhance the charge and discharge
capacity [51,58,60], reduce the temperature rise [89], and prolong
the lifetime [58,87–90] of LIBs in comparison to the equivalent
CCCV method. At the same time, charging the battery above the
manufacturer’s cut-off voltage has a negative impact on the battery
lifetime [51,72,91]. In addition, the cut-off voltage-based criteria
are simple and easily implementable compared to other criteria.
3.2. C-rate finding methods
The charging current is one of the most significant factors that
affects the useful life and performance of LIBs. The polarization
inside the LIB rises as the current level increases. As a result, the
higher the C-rate, generally, the shorter the cycle life of LIB. There
are numerous methods in the literature for determining and rec-
ommending the charging current at each stage. To determine the
charging current at each stage, researchers have employed analyt-
ical techniques, optimization methods, and empirical approaches.
The details of each method are discussed in the following.
3.2.1. Empirical approach
An empirical approach was employed in the literature to deter-
mine the optimal C-rate of each stage. However, the C-rate was
first chosen randomly, and the optimal C-rates were determined
through laboratory experiments. In ref. [58], the authors conducted
thirty-six experiments to find the optimal C-rate (0.6, 0.3, 0.15, and
0.1 C). Therefore, much experimentation is generally required
when using the empirical approach to determine the optimal C-
rate. This method is also not standardized yet. Table 3 lists a few
references that employed an empirical approach to determine
the C-rate to charge the LIBs. Fig. 5 illustrates the complete proce-
dure used to evaluate the LIB performance and lifetime, where N
indicates the specific number of cycles after which the capacity
and lifetime of LIB are analyzed. The conclusions from the numer-
ous articles [50,58,60,71] are diverse, where no similarities are
found. The authors used different transition criteria for switching
from one stage to the next one. Distinct C-rates have different
effects on charging time and performance at each stage. In Sec-
tion 4, the effect of different numbers of stages, different stage
transition criteria, and the C-rate will be discussed in more detail.
3.2.2. Analytical technique
An analytical technique can also determine the optimal C-rate
for charging the LIB with the MSCC strategy. The authors analyzed
the differential equations through circuit modeling and employed
the first-order derivative test to minimize the charging time
[51,75]. The optimal C-rate is determined numerically using Eq.
(4) and then validated through experiments [41,62].
I
n
¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
I
n1
I
nþ1
pn3ð4Þ
In this case, I
n
represents the n-stage current, with I
n1
repre-
senting the previous stage current and I
n+1
expressing the next
stage current. Eq. (4) is only utilized when the first and last stage
charging currents are predetermined by the manufacturer, and
the number of stages must be odd numbers.
3.2.3. Optimization methods
To determine the optimal charging current for LIBs, several
prior studies employed optimization methods, as indicated in
Table 4. In the literature, the authors utilized the Taguchi method
[30,47,76,82,86,89], particle swarm optimization [68,80,88],
genetic algorithm [73], ant colony system [87], numerical opti-
mization [92], and grey wolf optimization [52] for finding the opti-
mal charging pattern. Regarding the C-rate for each stage, the
literature provides rather diverse results. There are a variety of
optimization methods and the different number of stages used to
determine the ideal charging profile. With the objective of finding
the optimal charging current, the general problem formulation is
represented in the following Eqs. (5)–(10) [29].
Objective function:
min fxðÞ¼ I
1
;I
2
;:I
n
fg ð5Þ
Subject to:
t
c
t
max
ð6Þ
SOC
c
SOC
min
ð7Þ
T
c
tðÞ¼T
max
8t
t
0
;t
c
½ ð8Þ
D
T
c
tðÞ
D
T
max
8t
t
0
;t
c
½ ð9Þ
I
lb
I
j
I
ub
8j
1;n½ ð10Þ
The objective function changes depending on the application.
The main objective is to find the optimal charging pattern that
reduces charging time while providing a sufficient cycle life to LIBs
Table 2
Different numbers of charging stages used for charging transitions proposed by various authors.
Ref. Cell chemistry Multi-stage SOC-based charging (SOC in %)
S1 S2 S3 S4 S5 S6 S7 S8
[50] NCM 0%–70% 70%–80% 80%–100%
[74] Li-ion 0%–25% 25%–50% 50%–75% 75%–100%
[83] NMC 0%–20% 20%–40% 20%–60% 60%–80% 80%–100%
[80] LiFePO
4
10%–20% 20%–30% 30%–40% 40%–50% 50%–60% 60%–70% 70%–80% 80%–90%
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
235

for EV applications. Therefore, the objective function defined in Eq.
(5) is to find the lower currents at each stage while maintaining
good performance and lifetime. In Eq. (6),t
max
is adjusted to
increase charging currents and balance the objective function. Eq.
(6) is used to reduce the charging time (t
c
). Eq. (7) is incorporating
the minimum SOC charge (SOC
min
) into the MSCC charging strategy
(SOC
c
) during the charging. The thermal constraint is represented
via Eqs. (8) and (9).T
c
is the temperature during charging, and
D
Tshows the temperature change. The current constraint in Eq.
(10) includes both the upper (I
ub
) and lower (I
lb
) bounds of the cur-
rent range. The impact of the different optimization methods with
a distinct number of stages and the optimal C-rate on the charging
performance and LIB lifetime will be explored in greater depth in
Section 4.
4. Impact of MSCC charging on LIB
The benefits of the MSCC charging strategy have been demon-
strated by many researchers. However, the MSCC charging strategy
does not always have a positive influence on the performance and
lifetime of LIB. Therefore, it is necessary to explore the MSCC
charging strategy with a varying number of stages and distinct
transition conditions. The LIB performance is determined by four
parameters: charging time, charging efficiency, discharge capacity,
and temperature rise during charging. In the literature, some
researchers more exclusively concentrate on certain parameters.
The primary impact factor on the performance and lifetime is the
stage transition criteria and C-rate during the charging.
4.1. MSCC charging impact on performance
The impact of the MSCC charging strategy on the LIB perfor-
mance is evaluated based on the stage transition criteria. The
time-based transition is used in ref. [71] to analyze the impact of
different charging patterns and temperatures on the LIBs perfor-
mance. The optimal charging current pattern is found using the
empirical approach. The result indicates that the MSCC charging
strategy with time-based transition greatly reduced the charging
time and had a good capacity retention compared to constant cur-
rent (CC). The polarization effect is also small compared to the CC
charging method.
The voltage-threshold-based transition is another type of crite-
rion for stage transitions. In accordance with this criterion, various
voltage levels are established to shift the stages and implement the
optimal charging pattern. A numerical optimization technique was
employed to determine the optimal charging current [29]. The
technique implemented has a beneficial impact on the charging
time. The charging time is considerably shortened. In addition,
the threshold voltage has no detrimental effect on the polarization
and SEI growth since the LIB is below the cutoff voltage limit for
the majority of the time. The battery is not subject to any stress.
SOC-based transition is another method of stage transition. The
influence of the MSCC charging strategy with SOC-based transition
was investigated [50,68,74,80–82]. However, the optimal charging
current for each stage was determined using an empirical tech-
nique [50] and an optimization method [68,74,80–82]. The finding
indicated that SOC-based transitioning improves the performance
of LIBs. However, the implementation and estimation of SOC is dif-
ficult due to the nonlinear behavior of LIBs.
Table 3
Empirical approach for finding the optimal C-rate.
Ref. Cell chemistry Cell type Cell capacity No. of stages Criteria of transition between stages C-rate of each stage
[50] NCM Cylindrical 2.8 Ah Three SOC 1, 0.5, 0.3
[71] NCM Pouch - Three Time 1.8, 1.5, 0.9
[58] NCM Cylindrical 2 Ah Four Cut-off voltage 0.6, 0.3, 0.15, 0.1
[60] LiFePO
4
Cylindrical 1.37 Ah Ten Cut-off voltage 1, 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2, 0.1
Fig. 5. Flow chart for charging and discharging cycles for evaluation of MSCC
charging strategy and its impact on the LIB performance and lifetime.
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
236

The most often utilized criterion for stage transition is the cut-
off voltage. The cut-off voltage-based criterion was employed
[30,51,52,55,58,60,73,75,76,86–90,92,93]. However, different
authors employed distinct approaches to figure out the optimal
charging pattern. Compared to the CCCV technique, the results
indicate that charging time is shortened by 56.8% [88], charge
and discharge capacity is enhanced by 1.8% [51], the temperature
is lowered by 9.3 °C[76], and charging efficiency is improved by
2.8% [76]. The cut-off voltage-based criteria have the largest influ-
ence on the performance metrics.
Table 5 provides an overview of the impact of the MSCC charg-
ing technique on performance. The conclusion drawn from the lit-
erature is that, as claimed by numerous researchers, the MSCC
charging strategy reduces charging time and enhances charging
efficiency; both are important parameters. However, due to the
diversity in the methods for determining the optimal C-rate and
number of stages, there is some inconsistency in the literature
regarding the temperature rise during charging and charge/dis-
charge capacity compared to CCCV. Some authors observed an
increase in temperature and a decrease in charge/discharge capac-
ity when comparing the MSCC charging technique to the standard
CCCV method. The MSCC charging strategy has a positive impact
on charging time and charging efficiency but also has a negative
impact on temperature rise and charge/discharge capacity. Further
discussion on the impact of C-rate and variation in stages is dis-
cussed in Section 5.
4.2. MSCC charging impact on lifetime
To investigate the impact of the MSCC charging strategy on the
lifetime of LIB, various authors have employed different MSCC
charging techniques. The authors [92] combined a semi-empirical
aging model with an adaptive MSCC charging strategy. Different
scenarios were constructed to determine the optimal charging cur-
rent and number of stages. The results demonstrate that the bal-
ancing charging technique improves the lifetime of LIB by 3.6%
compared to the 0.5 C-CCCV charging method. In a similar manner,
the Taguchi method is utilized to figure out the optimal charging
pattern for a five-stage charging strategy [90]. A total number of
54 experiments were performed, and the results demonstrate that
the MSCC charging strategy provides 57% more cycle life in com-
parison to the CCCV method. Similarly, the same Taguchi method
is applied [89], and the result indicates that the acquired charging
pattern gives 60% more cycle life than the CCCV method. Grey wolf
optimization is used [52] to determine the optimal charging pat-
tern. The obtained results suggest a 79.6% improvement in cycle
life compared to the CCCV method. The particle swarm optimiza-
tion and the ant colony algorithm were employed to determine
the optimal charging pattern [88]. The results indicate a 21% and
25% improvement in life cycles compared to the CCCV method.
Thus, Table 6 provides a summary of the impact of the MSCC
charging strategy on the lifetime of LIBs. The conclusion from
Table 6 is that the MSCC charging strategy improves lifetime
regardless of the number of stages, cell chemistry, or transition cri-
teria. However, the majority of the researchers conducted their
investigations utilizing a five-stage charging strategy with a cut-
off voltage as the transition criterion.
5. Analysis and discussion
This section discusses how the key variables of various MSCC
charging strategies affect the lifetime and performance character-
istics (charging time, charge/discharge capacity, charging effi-
ciency, and temperature rise during charging) of LIBs. The
number of stages, the charging C-rate, and the criterion for stage
transitions are the primary impact factors of the MSCC charging
procedure. First, the impact factors of the main parameters are
assessed. The last part represents a discussion of the above
overview.
5.1. Impact of a variable number of stages
The literature has a wide range of results that can be used to
explore the impact factor of a different number of stages on life-
time and performance of LIB. The optimal numbers of stages and
LIB performance characteristics have been investigated [81].As
the number of charging stages increases from one to five, the
charging efficiency and charging time are improved significantly.
However, when the number of stages exceeds five, only marginal
improvement can be achieved [9,51,52,68,75,81,88,93]. Therefore,
it is not suggested to employ more than five stages since this
would increase the complexity of the charger control circuitry
without significantly enhancing the LIB performance and lifetime.
Table 4
Optimization methods for finding the optimal C-rate for MSCC charging strategy.
Ref. Cell
chemistry
Cell type Cell
capacity
No. of
stages
Optimization technique Criteria of transition
between stages
Optimal C-rate of each stage
[74] Li-ion Cylindrical 0.84 Ah Four Taguchi method SOC 1.4, 1, 0.7,0.4
[82] Li-Poly Pouch 5 Ah, 5.8
Ah
Four Taguchi method SOC 1.8, 1.3, 0.9, 0.5
[86] Li-ion Cylindrical 0.84 Ah Four Taguchi-based Particle swarm
optimization
Cut-off voltage 1.23, 1.08, 0.66, 0.26
[30] Li-ion Cylindrical 0.84 Ah Four Taguchi Cut-off voltage 1.4, 0.9, 0.75, 0.2
[73] Li-ion Cylindrical 3.5 Ah Four Genetic algorithm Cut-off voltage Starting from 1 C, then adaptive at
each stage
[87] Li-ion Cylindrical 0.93 Ah Five Ant-colony system Cut-off voltage 2.1, 1.7, 1.5,1.3, 1
[88] Li-ion Cylindrical 2.2 Ah Five Particle swarm Optimization Cut-off voltage 1.44, 1.18, 0.87, 0.71, 0.41
[76] Li-ion Cylindrical 3.15 Ah Five Taguchi method Cut-off voltage 1.55, 1, 0.6, 0.3, 0.2
[89] Li-ion - 0.6 Ah Five Taguchi method Cut-off voltage 1.5, 1.2, 0.9, 0.65, 0.4
[90] Li-ion - 0.65 Ah Five Taguchi method Cut-off voltage 1.45, 1.05, 1,0.7, 0.1
[52] Li-ion Cylindrical 2.5 Ah Five Grey wolf optimization Cut-off voltage 0.83, 0.47, 0.31, 0.21, 0.14
[80] LiFePO
4
Pouch 8 Ah Eight Particle swarm optimization SOC 4.9, 4.55, 3.76, 1.75, 1.63, 1.55, 1,
1.25
[29] NMC Cylindrical 3 Ah Ten Numerical optimization Threshold voltage 2.1, 2.37, 1.8, 1.6, 1.43, 1.1, 1.06,
0.73, 0.4, 0.3
[92] Li-ion Cylindrical 3.25 Ah Adaptive Particle swarm optimization Cut-off voltage Starting from 1 C, then adaptive at
each stage
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
237

Table 7 shows the relationship between the number of stages and
the performance parameters, lithium plating, and lifetime. Gener-
ally, the optimal MSCC charging strategy can prevent the precipita-
tion of lithium metal during the charging process [29,59,71]. Based
on the results in the literature, the five-stage charging strategy is
generally a suitable strategy compared to other stages in terms
of better performance parameters and lifetime.
5.2. Impact of different C-rate
The performance and lifetime of the LIBs are greatly impacted
by the charging pattern applied to the MSCC charging strategy.
Various authors have utilized different techniques to determine
the optimal C-rate, including empirical approaches [58], analytical
techniques [51], and optimization methods [52]. However, as the
Table 5
Impact of the MSCC charging strategy on the performance of LIBs (Y: Yes, N: No, S: Similar, -: no data available).
Ref Cell
chemistry
Cell type Cell
capacity
Compared
with
Transition
criteria
between
stages
No. of
stages
Method for
finding
optimal C-
rate
Performance parameters
Shortened
charging
time
Enhance
charge/
discharge
capacity
Temperature
rise
Improved
charging
efficiency
[71] NCM Pouch - CC Time Three Empirical Y Y - -
[29] NCM Cylindrical 3 Ah CCCV Threshold
voltage
Ten Optimization Y - - -
[50] NCM Cylindrical 2.8 Ah CCCV SOC Three Empirical Y Y N -
[81] Li-ion - 1 Ah CCCV SOC Four Optimization 18.25% - - -
[80] LiFePO
4
Pouch 8 Ah - SOC Eight Optimization Y - 4.1 °C-
[68] NMC Cylindrical 2.6 Ah CCCV SOC Five Optimization Y - N -
[82] Li-poly Pouch 5 Ah, 5.8
Ah
CCCV SOC Four Optimization Y N N Y
[74] Li-ion Cylindrical 0.84 Ah CCCV SOC Four Optimization 15.3% N N N
[83] Li-ion Coin cell 3.8 mAh CCCV SOC Five Empirical 20% - - -
[30] Li-ion Cylindrical 0.84 Ah CCCV Cut-off
voltage
Four Optimization Y N Y Y
[75] Li-ion Cylindrical 2.6 Ah CCCV Cut-off
voltage
Five Analytical 12% - - 0.54%
[60] LiFePO
4
Cylindrical 1.37 Ah CCCV Cut-off
voltage
Ten Empirical Y Y S -
[58] NCM Cylindrical 2 Ah CCCV Cut-off
voltage
Four Empirical Y Y S -
[86] Li-ion Cylindrical 0.84 Ah CCCV Cut-off
voltage
Four Optimization 43.7% - - -
[73] Li-ion Cylindrical 3.5 Ah CCCV Cut-off
voltage
Four Optimization Y - - N
[51] Li-ion Cylindrical 2.6 Ah CCCV Cut-off
voltage
Five Analytical 12% 1.8% N 0.54%
[87] Li-ion Cylindrical 0.93 Ah CCCV Cut-off
voltage
Five Optimization Y - - -
[88] Li-ion Cylindrical 2.2 Ah CCCV Cut-off
voltage
Five Optimization 56.80% N - 0.4%
[89] Li-ion - 0.6 Ah CCCV Cut-off
voltage
Five Optimization Y N N 1%
[90] Li-ion - 0.65 Ah CCCV Cut-off
voltage
Five Optimization 11.2% - - 1.02%
[92] Li-ion Cylindrical 3.25 Ah CCCV Cut-off
voltage
Adoptive Optimization 37% - - -
[55] LiFePO
4
Prismatic 20 Ah CCCV Cut-off
voltage
Five Optimization Y Y N -
[76] Li-ion Cylindrical 3.15 Ah CCCV Cut-off
voltage
Five Optimization Y N 9.3 °C 2.8%
[52] Li-ion Cylindrical 2.5 Ah CCCV Cut-off
voltage
Five Optimization 5.33% - N 0.48%
Table 6
Impact of the MSCC charging strategy on the lifetime of LIBs (Y: Yes).
Ref. Cell
chemistry
Cell type Cell
capacity
Method for finding optimal
C-rate
No. of
stages
Transition parameter between
stages
Compared
with
Improved
lifetime
[71] NCM Pouch - Empirical approach Three Time CC Y
[58] NCM Cylindrical 2 Ah Empirical approach Four Cut-off voltage CCCV Y
[94] LiFePO
4
Pouch 7 Ah Empirical approach Four SOC CCCV Y
[29] NMC Cylindrical 3 Ah Optimization method Ten Threshold voltage CCCV Y
[87] Li-ion Cylindrical 0.93 Ah Optimization method Five Cut-off voltage CCCV 25%
[88] Li-ion Cylindrical 2.2 Ah Optimization method Five Cut-off voltage CCCV 21%
[52] Li-ion Cylindrical 2.5 Ah Optimization method Five Cut-off voltage CCCV 79.6%
[89] Li-ion - 0.6 Ah Optimization method Five Cut-off voltage CCCV 60%
[90] Li-ion - 0.65 Ah Optimization method Five Cut-off voltage CCCV 57%
[92] Li-ion Cylindrical 3.25 Ah Optimization method Adaptive Cut-off voltage CCCV 3.6%
[83] Li-ion Coin Cell 3.8 mAh Empirical approach Five SOC CCCV Y
[59] NMC Cylindrical 2 Ah Empirical approach Five Threshold voltage CCCV Y
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
238

charging current increases, charging time decreases [95], the tem-
perature rises [95], and charging efficiency and cycle life decrease.
On the one hand, the high charging current, e.g., 2 C, results in low
energy efficiency, short charging time, and high-temperature rise.
On the other hand, the low charging current, e.g., 0.3 C, results in
high energy efficiency, long charging time, low-temperature rise,
and potentially improved lifetime [82]. Thus, the combination of
the various C-rates is required to balance all performance charac-
teristics and the lifetime of LIBs. Furthermore, the analytical
method would give a solid foundation for moving forward. For
determining the optimal charging profile, only the information of
the charging current levels of the first stage and last stage is
required. These current limits are usually provided by manufactur-
ers in terms of high charge current and nominal charge current.
The charging capacity usually depends upon the last stage of the
charging current. The charged capacity is inversely proportional
to the last stage charging current and charging time. As low as
the last stage charging current as much energy is stored in the
LIB. But the charging time would increase. However, finding the
optimal charge pattern using the 1-resistor-capacitor (RC) circuit
[51] is not as accurate as 2RC model. The optimal charging pattern
utilizing the 2RC model will be more accurate. The charging cur-
rent must also have an impact based on the cell chemistry.
5.3. Impact of the stage transition criteria
The stage transition criterion has an effect on the performance
and lifetime of the LIBs. As shown in Tables 5 and 6, the cut-off
voltage-based transition is widely employed because it enhanced
the performance and lifetime of LIBs. In the first stage, the voltage
rises to cut-off voltage rapidly compared to other criteria. In this
way, the cut-off voltage criteria increase the charging speed com-
pared to other transition criteria. Table 8 illustrates the impact fac-
tor of different stage transition criteria based on some articles
[50,52,59,71]. Generally, the most suitable criterion is the cut-off
voltage. In addition, the cut-off voltage conditions are easier to
implement in comparison to the SOC and threshold voltage criteria
[52]. According to the previous research, when cut-off voltage-
based criteria are utilized, the LIBs lifetime was extended by
79.6%, charging time was shortened by 5.33%, the temperature
dropped by 26%, and charging efficiency was increased by 0.48%
[52].
The following conclusions can be drawn from the aforemen-
tioned discussion and evaluation results.
From all of the impact factors, the optimal C-rate of the MSCC
charging strategy had the highest impact on the performance
and lifetime of LIBs. Beginning with a charging current of 1–
1.5 C, the lifetime is positively impacted; however, charging
currents beyond 2 C have a negative impact on the LIB lifetime,
discharge capacity, and charging efficiency.
High C-rate charging is associated with increased polarization
due to transport and kinetic overpotentials, which is favorable
for lithium plating.
Several factors influenced charging speed, including the number
of stages and the first and last stage charge current. Increasing
the number of stages improves capacity utilization and charging
efficiency.
The relationship between charging current and charging time is
almost linear. However, the cycle life decreases as the charging
current increases.
A higher charging current led to a larger temperature variation.
The rise in the temperature will have a negative impact on the
safety and lifetime of LIBs.
The overall impact of the MSCC charging strategy is positive on
the performance and lifetime. The main challenge is to balance
the charging speed and capacity utilization, where different fac-
tors will make the priority.
6. Conclusions
The MSCC charging techniques, i.e., three, four, five, eight, and
ten, are all summarized in this article. Different methods for deter-
mining the optimal C-rate were investigated based on various
MSCC charging strategies. An overview of MSCC charging method-
ologies and their impact on performance metrics and lifetime is
Table 7
Impact factor of the different number of stages on LIBs performance and lifetime (+: positive impact, N/A: no data available).
Ref. Strategy Compared
with
Shortened
charging time
Improved charging
efficiency
Enhanced charge/
discharge capacity
Reduction in
temperature rise
Improved
lifetime
Reduction in
lithium plating
[71] Three
stages
CC + N/A + N/A + +
[58] Four
stages
CCCV + N/A + + + N/A
[52] Five
stages
CCCV + + N/A + + N/A
[59] Five
stages
CCCV + N/A N/A N/A + +
[80] Eight
stages
CCCV + N/A N/A + N/A N/A
[29] Ten
stages
CCCV + N/A N/A N/A + +
Table 8
Impact factor of the different stage transition criteria on performance parameters and lifetime of LIBs (+: positive impact, N/A: no data available).
Ref. Stage
transition
criteria
Compared
with
Shortened
charging time
Improved
charging
efficiency
Enhanced charge/
discharge capacity
Reduction in
temperature rise
Improved
lifetime
Reduction in
lithium plating
[52] Cut-off voltage CCCV + + N/A + + N/A
[59] Threshold
voltage
CCCV + N/A N/A N/A + +
[50] SOC CCCV + N/A + + N/A N/A
[71] Time CCCV + N/A + N/A + +
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
239

described. The key impact factors are studied and analyzed. The
factors that had the most impact on both performance and lifetime
were charging current, the number of stages, and stage transition
condition. However, the primary factor that significantly affects
the charging time, charge/discharge capacity, temperature rise,
charging efficiency, and lifetime is the charging profile. The charg-
ing time is shortened as the charging current rises above a partic-
ular level, but at the same time, the charging efficiency and lifetime
deteriorate. The second factor is the number of charging stages
required for optimal performance. For a greater charge capacity
and longer lifecycles, the five stages are appropriate. The charging
efficiency and lifetime are somewhat enhanced by more than five
stages, but it makes the system complex. Lastly, the criteria in
between stages have a significant impact. The optimal criteria for
changing from one stage to the next stage are the cut-off voltage.
It is simple to implement and easy to control. Future research must
also look into how the MSCC charging approach affects LIB temper-
ature, charge/discharge capacity, lifetime, and speed in order to
design an enhanced fast-charging algorithm for EVs.
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
References
[1] IEA, 2022. https://www.iea.org/reports/global-ev-outlook-2022.
[2] J. Rogelj, O. Geden, A. Cowie, A. Reisinger, Nature 591 (2021) 365–1363.
[3] M.U. Tahir, M. Anees, H.A. Khan, I. Khan, N. Zaffar, T. Moaz, J. Energy Storage 36
(2021).
[4] X. Huang, Y. Li, A.B. Acharya, X. Sui, J. Meng, R. Teodorescu, D. Stroe, Energies
13 (2020) 2458.
[5] Z. Wang, G. Feng, D. Zhen, F. Gu, A. Ball, Energy Rep. 7 (2021) 5141–5161.
[6] M.U. Tahir, T. Moaz, H.A. Khan, N.A. Zaffar, I. Khan, in: IEEE Texas Power and
Energy Conference (TPEC) (2020) 1–6.
[7] M.U. Mutarraf, Y. Guan, L. Xu, C. Su, J.C. Vasquez, J.M. Guerrero, Sustain. Energy
Technol. Assessments 52 (2022).
[8] R. Xiong, Battery management algorithm for electric vehicles, Springer,
Singapore, 2020, 10.1007/978-981-15-0248-4.
[9] H. Min, W. Sun, X. Li, D. Guo, Y. Yu, T. Zhu, Z. Zhao, Energies 10 (2017) 709.
[10] M.A. Monem, K. Trad, N. Omar, O. Hegazy, B. Mantels, G. Mulder, P. Van den
Bossche, J. Van Mierlo, Appl. Energy 152 (2015) 143–155.
[11] P. Keil, A. Jossen, J. Energy Storage 6 (2016) 125–141.
[12] X. Han, L. Lu, Y. Zheng, X. Feng, Z. Li, J. Li, M. Ouyang, ETransportation 1 (2019).
[13] N. Omar, M.A. Monem, Y. Firouz, J. Salminen, J. Smekens, O. Hegazy, H.
Gaulous, G. Mulder, P. Van den Bossche, T. Coosemans, Appl. Energy 113
(2014) 1575–1585.
[14] C. Chen, F. Shang, M. Salameh, M. Krishnamurthy, in: 2018 IEEE Transportation
Electrification Conference and Expo (ITEC) (2018) 695–701.
[15] S.J. An, J. Li, C. Daniel, D. Mohanty, S. Nagpure, D.L. Wood III, Carbon 105 (2016)
52–76.
[16] M.F. Rodrigues, K. Kalaga, S.E. Trask, D.W. Dees, I.A. Shkrob, D.P. Abraham, J.
Electrochem. Soc. 166 (2019) A996.
[17] B. Horstmann, F. Single, A. Latz, Curr. Opinion Electrochem. 13 (2019) 61–69.
[18] T. Waldmann, B. Hogg, M. Wohlfahrt-Mehrens, J. Power Sources 384 (2018)
107–124.
[19] X. Lin, K. Khosravinia, X. Hu, J. Li, W. Lu, Prog. Energy Combustion Sci. 87
(2021).
[20] A. Adam, J. Wandt, E. Knobbe, G. Bauer, A. Kwade, J. Electrochem. Soc. 167
(2020).
[21] M. Dotoli, E. Milo, M. Giuliano, R. Rocca, C. Nervi, M. Baricco, M. Ercole, M.F.
Sgroi, Batteries 7 (2021) 46.
[22] I. Laresgoiti, S. Käbitz, M. Ecker, D.U. Sauer, J. Power Sources 300 (2015) 112–
122.
[23] M. Xu, Z. Zhang, X. Wang, L. Jia, L. Yang, Energy 80 (2015) 303–317.
[24] K. Xu, Chem. Rev. 104 (2004) 4303–4418.
[25] S. Goriparti, E. Miele, F. De Angelis, E. Di Fabrizio, R.P. Zaccaria, C. Capiglia, J.
Power Sources 257 (2014) 421–443.
[26] T.R. Jow, S.A. Delp, J.L. Allen, J. Jones, M.C. Smart, J. Electrochem. Soc. 165
(2018) A361.
[27] X. Li, D. Yu, S.B. Vilsen, D.I. Store, J. Energy Chem. 82 (2023) 103–121.
[28] N. Wassiliadis, J. Schneider, A. Frank, L. Wildfeuer, X. Lin, A. Jossen, M.
Lienkamp, J. Energy Storage 44 (2021).
[29] R. Mathieu, O. Briat, P. Gyan, J. Vinassa, J. Energy Storage 40 (2021).
[30] C. Lee, C. Hsu, S. Hsu, J. Jiang, IEEE Trans. Aerospace Electron. Syst. 57 (2021)
2704–2714.
[31] S.S. Zhang, ChemElectroChem 7 (2020) 3569–3577.
[32] W. Vermeer, M. Stecca, G. R. C. Mouli, P. Bauer, in: 2021 IEEE 19th International
Power Electronics and Motion Control Conference (PEMC) (2021), pp. 217–
224.
[33] C. Pastor-Fernández, K. Uddin, G.H. Chouchelamane, W.D. Widanage, J. Marco,
J. Power Sources 360 (2017) 301–318.
[34] X. Huang, W. Liu, A.B. Acharya, J. Meng, R. Teodorescu, D. Stroe, IEEE Trans. Ind.
Electron. 69 (2021) 10144–10153.
[35] C. Heubner, M. Schneider, A. Michaelis, Adv. Energy Mater. 10 (2020) 1902523.
[36] M. Weiss, R. Ruess, J. Kasnatscheew, Y. Levartovsky, N.R. Levy, P. Minnmann, L.
Stolz, T. Waldmann, M. Wohlfahrt-Mehrens, D. Aurbach, Adv. Energy Mater. 11
(2021) 2101126.
[37] C. Bolsinger, K.P. Birke, J. Energy Storage 21 (2019) 222–230.
[38] Y. Gao, J. Jiang, C. Zhang, W. Zhang, Z. Ma, Y. Jiang, J. Power Sources 356 (2017)
103–114.
[39] C. Chen, Z. Wei, A.C. Knoll, IEEE Trans. Trans. Electrification 8 (2021) 3068–
3089.
[40] M. Keyser, A. Pesaran, Q. Li, S. Santhanagopalan, K. Smith, E. Wood, S. Ahmed, I.
Bloom, E. Dufek, M. Shirk, J. Power Sources 367 (2017) 228–236.
[41] P.H. Notten, J.O. Het Veld, J. Van Beek, J. Power Sources 145 (2005) 89–94.
[42] A. Tomaszewska, Z. Chu, X. Feng, S. Okane, X. Liu, J. Chen, C. Ji, E. Endler, R. Li, L.
Liu, ETransportation 1 (2019) 100011.
[43] Y. Gao, X. Zhang, Q. Cheng, B. Guo, J. Yang, IEEE Access 7 (2019) 43511–43524.
[44] D.R.R. Kannan, M.H. Weatherspoon, J. Power Sources 479 (2020).
[45] J.M. Amanor-Boadu, A. Guiseppi-Elie, E. Sánchez-Sinencio, Energies 11 (2018)
2162.
[46] X. Huang, The Effects of Pulsed Charging Current on the Performance and
Lifetime of Lithium-Ion Batteries, Aalborg Universitetsforlag, 2021.
[47] J.M. Amanor-Boadu, A. Guiseppi-Elie, E. Sánchez-Sinencio, IEEE Trans. Ind.
Electron. 65 (2018) 8982–8992.
[48] J.M. Amanor-Boadu, M.A. Abouzied, E. Sánchez-Sinencio, IEEE Trans. Ind.
Electron. 65 (2018) 7383–7394.
[49] J. Li, E. Murphy, J. Winnick, P.A. Kohl, J. Power Sources 102 (2001) 302–309.
[50] D. Ji, L. Chen, T. Ma, J. Wang, S. Liu, X. Ma, F. Wang, J. Power Sources 437 (2019).
[51] A.B. Khan, W. Choi, IEEE Trans. Energy Convers. 33 (2018) 1132–1140.
[52] G. Chen, Y. Liu, S. Wang, Y. Luo, Z. Yang, J. Energy Storage 33 (2021).
[53] P. Makeen, H.A. Ghali, S. Memon, Future Transp. 2 (2022) 15.
[54] Y. Zhang, S. Xu, T. Wu, Global Energy Interconnection 5 (2022) 143–153.
[55] X. Wu, Y. Xia, J. Du, X. Gao, S. Nikolay, IEEE Trans. Transp Electrification (2022).
[56] M. Khalid, F. Ahmad, B.K. Panigrahi, L. Al-Fagih, J. Energy Storage 53 (2022).
[57] M. Brenna, F. Foiadelli, C. Leone, M. Longo, J. Electr. Eng. Technol. 15 (2020)
2539–2552.
[58] J. Liu, Q. Duan, H. Chen, J. Sun, Q. Wang, Sustain. Energy Fuel. 2 (2018) 1726–
1736.
[59] M. Dotoli, E. Milo, M. Giuliano, A. Tiozzo, M. Baricco, C. Nervi, M. Ercole, M.F.
Sgroi, Batteries 8 (2022) 88.
[60] X. Wu, C. Hu, J. Du, J. Sun, Math. Problems Eng. 2015 (2015).
[61] S. Zhu, C. Hu, Y. Xu, Y. Jin, J. Shui, J. Energy Chem. 46 (2020) 208–214.
[62] C.Z. El-Bayeh, K. Alzaareer, A.I. Aldaoudeyeh, B. Brahmi, M. Zellagui, World
Electr. Vehicle J. 12 (2021) 11.
[63] M. Li, M. Feng, D. Luo, Z. Chen, Cell Rep. Phys. Sci. 1 (2020).
[64] W. Xie, X. Liu, R. He, Y. Li, X. Gao, X. Li, Z. Peng, S. Feng, X. Feng, S. Yang, J.
Energy Storage 32 (2020).
[65] P. Bastida-Molina, E. Hurtado-Pérez, M.C.M. Gómez, C. Vargas-Salgado, Renew.
Energy 179 (2021) 737–755.
[66] J. Sieg, A.U. Schmid, L. Rau, A. Gesterkamp, M. Storch, B. Spier, K.P. Birke, D.U.
Sauer, Appl. Energy 305 (2022).
[67] P. Nambisan, P. Saha, M. Khanra, J. Energy Storage 41 (2021).
[68] J. Sun, Q. Ma, C. Tang, T. Wang, T. Jiang, Y. Tang, IEEE Trans. Veh. Technol. 69
(2020) 14141–14149.
[69] J. Liu, G. Lin, S. Huang, Y. Zhou, Y. Li, C. Rehtanz, IEEE Trans. Trans.
Electrification 7 (2020) 1112–1122.
[70] P.R. Chinnam, A.M. Colclasure, B. Chen, T.R. Tanim, E.J. Dufek, K. Smith, M.C.
Evans, A.R. Dunlop, S.E. Trask, B.J. Polzin, A.C.S. Appl, Energy Mater. 4 (2021)
9133–9143.
[71] F. An, R. Zhang, Z. Wei, P. Li, RSC Adv. 9 (2019) 21498–21506.
[72] K. Liu, X. Hu, Z. Yang, Y. Xie, S. Feng, Energy Convers. Manag. 195 (2019) 167–
179.
[73] Y. Li, K. Li, Y. Xie, B. Liu, J. Liu, J. Zheng, W. Li, J. Energy Storage 37 (2021).
[74] C. Lee, M. Chen, S. Hsu, J. Jiang, J. Energy Storage 18 (2018) 528–537.
[75] A.B. Khan, V. Pham, T. Nguyen, W. Choi, in: 2016 IEEE Transportation
Electrification Conference and Expo, Asia-Pacific (ITEC Asia-Pacific), 2016,
pp. 381–385.
[76] L. Jiang, Y. Li, Y. Huang, J. Yu, X. Qiao, Y. Wang, C. Huang, Y. Cao, Appl. Energy
259 (2020).
[77] J. Lin, X. Liu, S. Li, C. Zhang, S. Yang, Int. J. Heat Mass Transfer 167 (2021).
[78] Y. Zeng, D. Chalise, S.D. Lubner, S. Kaur, R.S. Prasher, Energy Storage Mater. 41
(2021) 264–288.
[79] Y. Liu, Y. Zhu, Y. Cui, Nat. Energy 4 (2019) 540–550.
[80] J. Sun, Q. Ma, R. Liu, T. Wang, C. Tang, Int. J. Energy Res. 43 (2019) 7672–7681.
[81] L. Dung, J. Yen, 2010 IEEE International Symposium on Industrial Electronics,
2010, pp. 2286–2291.
[82] T.T. Vo, X. Chen, W. Shen, A. Kapoor, J. Power Sources 273 (2015) 413–422.
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
240

[83] Y. Li, J. Guo, K. Pedersen, L. Gurevich, D. Stroe, J. Energy Chem. 80 (2023) 237–
246.
[84] J. Meng, M. Ricco, G. Luo, M. Swierczynski, D. Stroe, A. Stroe, R. Teodorescu,
IEEE Trans. Ind. Appl. 54 (2017) 1583–1591.
[85] J.P. Rivera-Barrera, N. Muñoz-Galeano, H.O. Sarmiento-Maldonado, Electronics
6 (2017) 102.
[86] C. Lee, T. Chang, S. Hsu, J. Jiang, J. Energy Storage 21 (2019) 301–309.
[87] Y. Liu, J. Teng, Y. Lin, IEEE Trans. Ind. Electron. 52 (2005) 1328–1336.
[88] S. Wang, Y. Liu, IEEE Trans. Ind. Electron. 62 (2014) 2983–2993.
[89] Y. Liu, Y. Luo, IEEE Trans. Ind. Electron. 57 (2009) 3963–3971.
[90] Y. Liu, C. Hsieh, Y. Luo, IEEE Trans. Energy Convers. 26 (2011) 654–661.
[91] B. Wang, H. Min, W. Sun, Y. Yu, Energies 14 (2021) 1776.
[92] Y. Li, K. Li, Y. Xie, J. Liu, C. Fu, B. Liu, Renew. Energy 146 (2020) 2688–2699.
[93] Y. Wang, G. Zhao, C. Zhou, M. Li, Z. Chen, IEEE Trans. Trans Electrification
(2022).
[94] M. Abdel-Monem, K. Trad, N. Omar, O. Hegazy, P. Van den Bossche, J. Van
Mierlo, Energy 120 (2017) 179–191.
[95] B. Lu, Y. Zhao, Y. Song, J. Zhang, Electrochim. Acta 288 (2018) 144–152.
M. Usman Tahir, A. Sangwongwanich, D.-I. Stroe et al. Journal of Energy Chemistry 84 (2023) 228–241
241