Original Paper

Measurement and Control

2019, Vol. 52(3-4) 193–201
An equivalent circuit model analysis Ó The Author(s) 2019
Article reuse guidelines:
for the lithium-ion battery pack in pure sagepub.com/journals-permissions
DOI: 10.1177/0020294019827338

electric vehicles journals.sagepub.com/home/mac

Jie Su , Maosong Lin, Shunli Wang, Jin Li, James Coffie-Ken and Fei Xie

Abstract
According to the demand of vehicle lithium-ion battery pack, the splice equivalent circuit model is constructed. First, a
joint experiment of intermittent discharge and hybrid power pulse characterization, basis of the requirements of para-
meter identification for the model, is designed to identify the parameters. Study shows that it can identify the para-
meters. Second, the splice equivalent circuit model of vehicle lithium-ion battery pack is simulated by MATLAB/Simulink,
which shows the model is feasible to describe the vehicle lithium-ion battery pack. The simulation error of ohmic inter-
nal resistance R0 should be less than 0.05 mO. Study suggests that the vehicle lithium-ion battery pack has a stable dis-
charge period within the state-of-charge range of [20%, 80%]. However, when stage of charge is below 20%, vehicle
lithium-ion battery pack is no longer stable and the parameters of the splice equivalent circuit model change
dramatically.

Keywords
Lithium-ion battery pack, pure electric vehicle, splice equivalent circuit model, hybrid power pulse characterization test,
parameter identification

Date received: 17 October 2018; accepted: 22 December 2018

Introduction is constructed, and the structure of the model is rela-

tively simple. Based on the internal resistance model,
Pure electric vehicles as clean vehicles have been paid the Thevenin model6–11 is constructed by introducing
enough attention because of the pressure of conven- the resistor and capacitor parallel circuits to describe
tional power vehicles on fossil energy consumption.1 the polarization effect inside the battery packs. The
Lithium-ion battery pack is the energy source for pure dynamic characteristics of the battery packs are simu-
electric vehicle. Its equivalent circuit modeling methods lated to some extent. Comparing the Thevenin model,
have attracted attention very much.2 The equivalent cir- the PNGV model12,13 considers the cumulative effect of
cuit modeling of pure electric vehicle lithium-ion bat- load current on lithium-ion battery packs. At the same
tery pack is based on experimental data without time, in order to reflect accurately the changing rule of
considering the internal chemical reactions. The equiva- battery, researchers have successively developed various
lent circuit model is used to describe the dynamic char- resistance-capacitance (RC) equivalent circuit mod-
acteristics of lithium-ion battery pack. It has strong els.14–24 More recent study in Wu et al.25 fully considers
applicability. the difference of internal resistance between charging/
In the past researches, the modeling of lithium-ion
discharging and then improves the Thevenin model. An
battery pack has been studied continuously and deeply.
Shepherd proposed a Shepherd model to describe
the electrochemical characteristics of batteries directly School of Information Engineering, Engineering and Technology Center,
by voltage and current for the first time.3,4 Then the Southwest University of Science and Technology, Mianyang, China
researchers built the Unnewehr model by analyzing the
reaction process. After that, Nernst proposed the Corresponding author:
Shunli Wang, School of Information Engineering, Engineering and
Nernst model,5,6 which improved the description of Technology Center, Southwest University of Science and Technology,
non-linearity. While considering the working character- Mianyang 621010, China.
istics of the battery packs, the internal resistance model Email: [email protected]

Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License
(http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without
further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/
open-access-at-sage).
194 Measurement and Control 52(3-4)

Figure 1. S-ECM model for vehicle lithium-ion pack.

nRC equivalent circuit model is proposed in Lai et al.26 resistance Rs is connected in parallel to characterize the
The high-order RC model is robust to the variations of self-discharge effect of the lithium-ion battery pack. Ro
model parameters and sensor errors. Wang et al.27 pro- is the benchmark ohmic internal resistance of the bat-
posed a splice equivalent circuit model (S-ECM) for the tery pack, which is used to describe the ohmic effect of
aerial lithium-ion battery pack, which can be used for the lithium-ion battery pack. Rd is a small variable of
the accurate estimation of state of charge (SoC). ohmic internal resistance caused by discharging. Rc is a
However, the difference between the monomers of small variable of ohmic internal resistance caused by
lithium-ion battery pack was not considered during the charging. They can reflect the dynamic characteristics
equivalent circuit modeling process in most of the of lithium-ion battery pack during discharging/charg-
above-mentioned researches. Disadvantages still exist ing. Rp is used to characterize the polarization resis-
in some of the simulated description of lithium-ion bat- tance of lithium-ion battery pack, and Cp is the
tery packs. In order to solve these issues, this paper polarization capacitance of lithium-ion battery pack. Rp
study the equivalent circuit model based on the first- and Cp constitute a first-order RC parallel circuit to
order RC model. First, the time-varying voltage source represent the polarization effect of lithium-ion battery
and the time-varying resistance are connected in series. pack and simulate the relaxation effect of lithium-ion
They are used to describe the effects of the equilibrium battery pack, thus realizing the transient response
state between monomers on open-circuit voltage (OCV) description of lithium-ion battery pack. Ud and Rd are
and ohmic internal resistance, respectively. Then, the used to describe the equilibrium state between internal
ohmic internal resistance is divided into benchmark interconnect monomers. They reflect the effects of the
ohmic internal resistance and charge/discharge differen- equilibrium state between monomers on OCV and
tial ohmic internal resistance to simulate the change ohmic internal resistance, respectively. UL is the termi-
characteristics of ohmic internal resistance of charging/ nal voltage of lithium-ion battery pack for pure electric
discharging. Meanwhile, the self-discharge internal vehicle, and I(t) are the current value at both ends.
resistance is used to simulate the self-discharge effect of
vehicle battery packs. After that, this paper discusses
one method to identify these parameters. Finally, this Parameter identification
model is applied to vehicle lithium-ion battery packs. In order to obtain model parameters, the intermittent
The rest of the paper is organized as follows. In sec-
discharge and hybrid power pulse characterization
tion ‘‘Theoretical analysis,’’ the construction method of
(HPPC) joint experiments of vehicle lithium-ion battery
equivalent circuit model and parameter identification
pace are designed. The process of designing HPPC is
method are discussed. In section ‘‘Experimental analy-
shown in Figure 2.
sis,’’ the proposed model is verified experimentally.
According to Figure 2, the parameters of lithium-ion
battery pack for pure electric vehicle can be obtained.
Theoretical analysis The counting method is as follows:

Construction of equivalent circuit model 1. The calculation of Uoc: When the lithium-ion bat-
According to the research needs of the pure electric tery pack is fully shelved, the internal electrochemi-
vehicle power battery packs, with existing research, cal reaction is stable, which is considered Uoc = UL
considering ohmic effect, polarization effect, and self- at this time. Therefore, in Figure 2, Uoc = U1.
discharge effect of the battery packs, the models-ECM 2. Calculate Ro: Ro describes the instantaneous vol-
is constructed. tage drop from U1 to U2 in Figure 2. Therefore
Figure 1 is the S-ECM model of vehicle lithium-ion
battery pack. Uoc is the OCV of the battery pack, which
represents the change of the terminal voltage OCV of U1  U2
R0 = ð1Þ
the vehicle lithium-ion battery pack, and a large I
Su et al. 195

,
Table 1. Experimental requirements.
8
8 Parameter Experimental requirements

8 Uoc SoC = 1 starts intermittent discharge and

8 shelving measurement
Rs Measuring OCV at present and after 30 days
,$

8

89
8
Ro HPPC test28,29 in intermittent discharge
8 Rp HPPC test28,29 in intermittent discharge
8 Cp HPPC test28,29 in intermittent discharge
Rd Calculation of pulse discharge combined with
8
ohmic resistance
Rc Calculation of pulse charge combined with ohmic
W W W W resistance
WV
Ud Combining the evaluation results of equilibrium
Figure 2. HPPC testing process. states to get function
Rd Combining the evaluation results of equilibrium
states to get function
In formula (1), I is the current value of 1C5A pulse
discharge. SoC: state of charge; OCV: open-circuit voltage; HPPC: hybrid power
pulse characterization.
3. The calculation of Rd and Rc: The ohmic internal
resistance of vehicle lithium-ion batteries will cause
the terminal voltage to rise and fall instantaneously 6. The calculation of Ud and Rd: Ud and Rd are used
during pulse charging and discharging. However, to describe the equilibrium state between the mono-
there are differences in ohmic resistance to charg- mer of pure vehicle lithium-ion battery pack. Their
ing and discharging. The difference is expressed by calculation methods are described in formula (5)
Rd and Rc. Their formulas are as follows
 8
Rd = U3 U 4
 R0 >
> Ud = e UOC
I
U7 U8 ð2Þ >
> Rd = e R0
Rc = I  R0 >
>
< n h
P i2
1 Uci EðUc Þ
e= n EðUc Þ ð5Þ
where U3, U4, U7, and U8 in formula (2) are shown in >
> i=1
Figure 2, and I is described in formula (1). >
> P
n
>
> ð 1
: E U cÞ = n Uci
i=1
4. The extraction of Rs: For the sake of obtaining Rs,
CC-CV is used to charge the pure vehicle lithium- In formula (5), n denotes the number of lithium-ion
ion battery packs to make SoC = 1. Then the bat- battery single monomer in the pure vehicle lithium-ion
tery is kept on hold for 1 h and the OCV of the bat- battery pack and Uci denotes the voltage value of ith
tery pack UOCRs is measured. The SoC value is lithium-ion battery pack single monomer only. e is the
SoCRs at this moment. Shelving for 30 days, the equilibrium state between the monomers, which is used
OCV of battery pack UOCRs2 is measured. The SoC to describe the degree of voltage discrepancy between
value is SoCRs2 at this moment. It is considered monomers.
that the whole self-discharge process is constant
current discharge and the current is Is. Assume that
the nominal capacity of the battery pack is Qn.
Experimental analysis
Accordingly, the calculation method of Rs can be Experimental requirements
obtained from formula (3)
There are nine parameters of the S-ECM model which
need to be identified. The identification requirements of
(
Rs = ðUOCRs2 UOCRs Þ these parameters are not identical exactly. The experi-
Is
DQ 3600ðSoCRs2 SoCRs ÞQn
ð3Þ mental requirements for identification are displayed in
Is = Dt = Dt Table 1.
5. Calculation of Rp and Cp: As shown in the HPPC
test shown in Figure 2, the formula for calculating Experimental design
the RC response parameter is as follows
To verify the correctness of the model and the accuracy
of parameter identification, the parameters of lithium
8
>
> Rp = U5 U
I
4 cobalt oxide battery commonly used in pure electric
< Cp = Rtp vehicles is identified by the power battery test system
ð4Þ (company: MACCOR; model: S4000; voltage accuracy:
> t =  t3 t2 
>
: U U
ln 1 5 4 0.02%; current accuracy: 0.02%). The parameters of
U1 U4
lithium-ion battery in this paper are shown in Table 2.
196 Measurement and Control 52(3-4)

Table 2. Battery parameters. Starting from SoC = 1, HPPC test is conducted every
5% SoC.
Project Parameter

Nominal voltage (V) 3.7

Discharge cut-off voltage (V) 3.2
Experimental result
Maximum loading voltage (V) 4.2 Experiments were carried out as in section
Self-discharge rate \ 5% (100% SoC, 28 days) ‘‘Experimental requirements’’. The relationship between
Rated capacity (A h) 30
S-ECM parameters and SoC of lithium-ion battery
SoC: state of charge. pack can be obtained by calculating the parameters
using the formulas provided in section ‘‘Parameter
identification.’’
The relation curve between UOC and SoC of vehicle
According to the requirements of Table 1, the inter- lithium-ion battery pack is shown in Figure 4.
mittent discharge and HPPC joint experiments of vehi- It is found that UOC and SoC have nonlinear positive
cle lithium-ion battery pack are designed. First, CC- correlations. When the SoC is above 80%, the UOC of
CV30 charging of vehicle lithium-ion battery pack will the lithium-ion battery pack decreases rapidly with the
enable SoC = 1. Then the battery is shelved for 40 min- discharge of the battery pack. Then it entered a stable
utes to stabilize the electrochemical reaction inside the period. When SoC ranged from 80% to 20%, the UOC
lithium cobalt oxide battery. Then, starting from decreased slowly with the battery discharged. When the
SoC = 1, 1C5A intermittent discharge is carried out, SoC is below 20%, the electrochemical reaction inside
and after each 5% SoC discharging, it was kept stable the battery pack is intense, and the UOC of the battery
for 40 minutes until SoC = 0. The intermittent dis- pack drops rapidly during the withdrawal period.
charge process is shown in Figure 3. Graph of relation between ohmic resistance R0 and
In Figure 3, a 3-min discharge 1C5A to discharge SoC is shown in Figure 5.
5% SoC on time is performed at an interval of R0 has a nonlinear negative correlation with SoC.
40 minutes, and an HPPC test is performed at the end With the discharging of the battery pack, the concen-
of the 40-min shelf. Aiming at the parameter identifica- tration of charged ions in the battery pack decreases,
tion of S-ECM model of pure vehicle battery pack, the and the ohmic resistance R0 increases. Ohmic resistance
constant current pulse charging and discharge experi- increases sharply when the battery is discharged at a
ments of 1C5A was carried out at 25 °C. depth of SoC less than 20%.
In Figure 2, according to classical HPPC steps, the In Figure 6, the black curve is the graph of relation
single HPPC experiment29 steps are as follows. The first between discharge resistance Rd and SoC, and the red
step is to carry out the 1C5A rate constant current pulse curve is the graph of relation between charge resistance
discharge 10 s, as shown in the t1 to t2 time section. The Rc and SoC. When SoC is higher than 20%, Rd and Rc
second step is to shelve for 40 s, as shown in the t2 to t3 fluctuate in a small range, and S-ECM introduces these
section of the diagram. The third step is to charge for two parameters to compensate for the inaccurate char-
10 s with 1C5A constant current pulse, as shown in the acterization of ohmic internal resistance on the internal
t3 to t4 time section. The fourth step is to shelve for 40 s. resistance of lithium-ion battery pack during charging




,&$









     

WV

Figure 3. Experimental current diagram.

Su et al. 197

8RF







8RF9 










     
62&

Figure 4. Graph of relation between UOC and SOC.

R0
2.5

2.0
R0 (mΩ)

1.5

1.0

0.5

0.0 0.2 0.4 0.6 0.8 1.0

SOC

Figure 5. Graph of relation between R0 and SoC.

and discharging. By comparing the two curves, it is fluctuation is small. When SoC is below 20%, Cp
found that the values of Rd and Rc are identical basi- increases rapidly, far beyond 3000 F. After that, Ud and
cally at the same SoC value. When SoC is below 20%, Rd are analyzed in Table 3.
both Rd and Rc increase rapidly. From Table 3, it can be found that the seven bat-
Figure 7 shows that Rc and SoC are not significantly teries used in this paper have good consistency at the
correlated; for obtaining the relationship between Rp beginning of discharge. As deep discharge proceeds, the
and SoC, we can continue the least square fitting. battery pack becomes inconsistent slowly. When SoC is
When the SoC is above 20%, the polarization capaci- less than 15%, the inconsistency increases rapidly, Ud
tance Cp stays between 2000 and 3000 F, and the and Rd increase accordingly. Therefore, Ud and Rd can
198 Measurement and Control 52(3-4)

5G
5F


5P¡








     

62&

Figure 6. Graph of relation between Rc , Rd, and SoC.

Table 3. Analysis of Ud and Rd.

SoC Monomer voltages (V) Ud (mV) Rd (mO)

Volt 1 Volt 2 Volt 3 Volt 4 Volt 5 Volt 6 Volt 7

100% 4.1292 4.1312 4.1313 4.1303 4.1283 4.1301 4.1337 0.0044 0.0000
95% 4.0864 4.0880 4.0877 4.0871 4.0862 4.0868 4.0904 0.0030 0.0000
90% 4.0462 4.0484 4.0472 4.0467 4.0467 4.0464 4.0508 0.0040 0.0000
85% 4.0053 4.0065 4.0048 4.0048 4.0056 4.0034 4.0080 0.0031 0.0000
80% 3.9630 3.9611 3.9578 3.9594 3.9626 3.9568 3.9616 0.0087 0.0000
75% 3.9245 3.9233 3.9201 3.9211 3.9242 3.9193 3.9237 0.0068 0.0000
70% 3.8884 3.8893 3.8870 3.8875 3.8891 3.8868 3.8899 0.0023 0.0000
65% 3.8619 3.8633 3.8618 3.8621 3.8632 3.8609 3.8642 0.0020 0.0000
60% 3.8402 3.8417 3.8405 3.8408 3.8416 3.8399 3.8429 0.0017 0.0000
55% 3.8222 3.8239 3.8224 3.8227 3.8235 3.8220 3.8246 0.0015 0.0000
50% 3.8083 3.8099 3.8085 3.8086 3.8097 3.8082 3.8104 0.0013 0.0000
45% 3.7945 3.7947 3.7932 3.7941 3.7953 3.7928 3.7948 0.0013 0.0000
40% 3.7797 3.7765 3.7737 3.7755 3.7783 3.7726 3.7761 0.0095 0.0000
35% 3.7600 3.7563 3.7537 3.7552 3.7583 3.7531 3.7558 0.0095 0.0000
30% 3.7465 3.7426 3.7405 3.7419 3.7446 3.7391 3.7426 0.0097 0.0000
25% 3.7272 3.7220 3.7197 3.7214 3.7245 3.7185 3.7218 0.0136 0.0000
20% 3.7053 3.6987 3.6967 3.6986 3.7015 3.6956 3.6986 0.0167 0.0000
15% 3.6913 3.6893 3.6870 3.6887 3.6902 3.6848 3.6896 0.0075 0.0000
10% 3.6257 3.5822 3.5639 3.5762 3.6035 3.5494 3.5813 1.0577 0.0005
5% 3.4016 3.2860 3.2266 3.2652 3.3453 3.1688 3.2838 10.5110 0.0062

SoC: state of charge.

describe the inconsistency between monomers. It also experimental data onto section ‘‘Experimental result’’ is
could quantitatively analyze their effects on OCV and imported into the simulation circuit and the ohmic
ohmic internal resistance. internal resistance R0 is taken as the observation. The
simulation results are shown in Figure 8.
It can be seen from Figure 8 that the simulation
Simulation and error analysis curve is highly consistent with the real value curve. For
The S-ECM shown in Figure 1 is simulated in the sake of quantitatively analyzing the simulation
MATLAB/Simulink to verify whether it can be used to results, the estimated error is calculated, as shown in
describe vehicle lithium-ion battery pack. The Figure 9.
Su et al. 199

Rp
20 Cp

19

8000
18

17

6000
Rp(mΩ)

Cp (F)
16

15

14 4000

13

12 2000

0.0 0.5 1.0

SOC

Figure 7. Graph of relation between Rp and Cp.

UHDO
VLPXODWLRQ



5P¡







     

62&

Figure 8. R0 simulation curve.

HUURU



(UURUP¡








     
62&

Figure 9. R0 simulation error diagram.

200 Measurement and Control 52(3-4)

Figure 9 is the R0 simulation error curve obtained References

from Figure 8, the simulation error = experimental 1. Ouyang MG, Du JY, Peng HE, et al. Progress review of
value 2 simulation value. Figure 9 shows that the error US-China joint research on advanced technologies for
of S-ECM of vehicle lithium-ion battery packs is less plug-in electric vehicles. Sci China Technol Sc 2018;
than 0.05 mO, which meets the description requirement 61(10): 1431–1445.
of vehicle lithium-ion battery pack. 2. Zhang RF, Xia BZ, Li BH, et al. State of the art of
lithium-ion battery SoC estimation for electrical vehicles.
Energies 2018; 11(7): 1820.
Conclusion 3. Shepherd CM. Design of primary and secondary cells II:
an equation describing battery discharge. J Electrochem
Soc 1965; 112: 657–664.
1. The S-ECM is feasible to describe pure electric 4. Wu Chunfeng ZT and Song K. Application of C.M
Shepherd battery model in modeling and simulation of
vehicle lithium-ion battery pack. The S-ECM
pure electric vehicle. Chinese J Mech 2017; 23(5): 47–52.
model of pure electric vehicle lithium-ion battery
5. Plett GL. Extended Kalman filtering for battery manage-
is simulated in MATLAB/Simulink. The simula- ment systems of LiPB-based HEV battery pack. J Power
tion error of ohmic internal resistance R0 is less Sources 2004; 134(2): 252–261.
than 0.05 mO. 6. Ma Y, Duan P, Sun Y, et al. Equalization of lithium-ion
2. It is feasible to identify the parameters of S- battery pack based on fuzzy logic control in electric vehi-
ECM model of pure electric vehicle lithium-ion cle. IEEE T Ind Electron 2018; 65(8): 6762–6771.
battery pack by the joint experiment of intermit- 7. Luo MJ, Guo YZ, Kang JQ, et al. Ternary-material
tent discharge of lithium-ion battery pack and lithium-ion battery SoC estimation under various ambi-
HPPC. This joint experiment can identify the ent temperature. Ionics 2018; 24(7): 1907–1917.
parameters of the S-ECM model. 8. Liu Z, Dang XJ and Sun HX. Online state of charge esti-
3. The pure electric vehicle lithium-ion battery mation for lithium-ion battery by combining incremental
pack has a stable discharge period within the autoregressive and moving average modeling with adap-
tive H-infinity filter. Math Probl Eng 2018; 2018:
SoC range of [20%, 80%]. The parameters of S-
7480602.
ECM model are stable and have little fluctua- 9. Lee KT, Dai MJ and Chuang CC. Temperature-compen-
tion. When SoC is lower than 20%, the concen- sated model for lithium-ion polymer batteries with
tration of charged ions in lithium-ion battery extended Kalman filter state-of-charge estimation for an
pack decreases, the electrochemical reaction implantable charger. IEEE T Ind Electron 2018; 65(1):
remains unstable, and the parameters of S-ECM 589–596.
model change sharply. 10. Tian JP, Xiong R and Yu QQ. Fractional-order model-
based incremental capacity analysis for degradation state
recognition of lithium-ion batteries. IEEE T Ind Electron
Declaration of conflicting interests 2019; 66(2): 1576–1584.
The author(s) declared no potential conflicts of interest with 11. Wei ZB, Zhao JY, Zou CF, et al. Comparative study of
respect to the research, authorship, and/or publication of this methods for integrated model identification and state of
article. charge estimation of lithium-ion battery. J Power Sources
2018; 402: 189–197.
12. Zhang ZL, Cheng X, Lu ZY, et al. SoC estimation of
Funding lithium-ion battery pack considering balancing current.
This research was supported by National Natural Science IEEE T Power Electr 2018; 33(3): 2216–2226.
Foundation of China (No. 61801407), National Key Specialty 13. Wang WD, Wang XT, Xiang CL, Wei C and Zhao YL.
Construction Project of Clinical Pharmacy (30305030698), Unscented Kalman filter-based battery SoC estimation
National Training Program of Innovation and Entrepreneurship and peak power prediction method for power distribution
for Undergraduates (No. 201810619018, No.201806190-27, of hybrid electric vehicles. IEEE Access 2018; 6: 35957–
No.201810619067), Southwest University of Science and 35965.
Technology student innovation fund precise funding project (No. 14. Shen YQ. A chaos genetic algorithm based extended Kal-
jz18-013).Sichuan Province Science and Technology Support man filter for the available capacity evaluation of lithium-
Program (No. 19ZDYF1098, 2018GZ0390, 2017FZ0013), ion batteries. Electrochim Acta 2018; 264: 400–409.
Scientific Research Fund of Sichuan (No. 17ZB0453), Teaching 15. Shen P, Ouyang MG, Lu LG, et al. The co-estimation of
Research Project (18lzx665, 18gjzx11, 18xnsu12), Sichuan state of charge, state of health, and state of function for
Science and Technology Innovation Cultivation Project (No. lithium-ion batteries in electric vehicles. IEEE T Veh
201810619078, 201810619096), RGU. Thanks to the sponsors. Technol 2018; 67(1): 92–103.
16. Rivera-Barrera JP, Munoz-Galeano N and Sarmiento-
Maldonado HO. SoC estimation for lithium-ion batteries:
ORCID iDs
review and future challenges. Electronics 2017; 6(4): 102.
Jie Su https://orcid.org/0000-0002-8043-2609 17. Wang SL, Shang LP, Li ZF, et al. Online dynamic equal-
James Coffie-Ken https://orcid.org/0000-0001-6659-1358 ization adjustment of high-power lithium-ion battery
Su et al. 201

packs based on the state of balance estimation. Appl battery pack based on the real-time detection and filter-
Energ 2016; 166: 44–58. ing. J Clean Prod 2018; 185: 187–197.
18. Mathew M, Janhunen S, Rashid M, et al. Comparative 25. Wu J, Xing XW, Liu XT, et al. Energy management strat-
analysis of lithium-ion battery resistance estimation tech- egy for grid-tied microgrids considering the energy storage
niques for battery management systems. Energies 2018; efficiency. IEEE T Ind Electron 2018; 65(12): 9539–9549.
11(6): 1490. 26. Lai X, Zheng YJ and Sun T. A comparative study of dif-
19. Chen J, Ouyang Q, Xu CF, et al. Neural network-based ferent equivalent circuit models for estimating state-of-
state of charge observer design for lithium-ion batteries. charge of lithium-ion batteries. Electrochim Acta 2018;
IEEE T Contr Syst T 2018; 26(1): 313–320. 259: 566–577.
20. Afshari HH, Attari M, Ahmed R, et al. Reliable state of 27. Wang SL, Fernandez C, Liu XH, et al. The parameter
charge and state of health estimation using the smooth identification method study of the Splice equivalent cir-
variable structure filter. Control Eng Pract 2018; 77: 1–14. cuit model for the aerial lithium-ion battery pack. Meas
21. Lai X, Qin C, Gao WK, et al. A state of charge estimator Control 2018; 51(5–6): 125–137.
based extended Kalman filter using an electrochemistry- 28. Yu QQ, Xiong R, Wang LY, et al. A comparative study
based equivalent circuit model for lithium-ion batteries. on open circuit voltage models for lithium-ion batteries.
Appl Sci 2018; 8(9): 1592. Chin J Mech Eng 2018; 31(1): 65.
22. Zhao LH, Ji GH and Liu ZY. Design and experiment of 29. INEE Laboratory. FreedomCAR battery test manual for
nonlinear observer with adaptive gains for battery state power-assist hybrid electric vehicles. U.S. Department of
of charge estimation. Energies 2017; 10(12): 2046. Energy, 2003, pp. 3–6, https://avt.inl.gov/sites/default/
23. Wei JW, Dong GZ, Chen ZH, et al. System state estima- files/pdf/battery/freedomcar_manual_04_15_03.pdf
tion and optimal energy control framework for multicell 30. Barai A, Uddin K, Widanage WD, et al. A study of the
lithium-ion battery system. Appl Energ 2017; 187: 37–49. influence of measurement timescale on internal resistance
24. Wang SL, Fernandez C, Chen MJ, et al. A novel safety characterisation methodologies for lithium-ion cells. Sci
anticipation estimation method for the aerial lithium-ion Rep 2018; 8: 21.