batteries

Article
SOC Estimation Methods for Lithium-Ion Batteries without
Current Monitoring
Zhaowei Zhang 1 , Junya Shao 1 , Junfu Li 1, *, Yaxuan Wang 2 and Zhenbo Wang 2

1 School of Automotive Engineering, Harbin Institute of Technology, Weihai 264209, China;

2200180129@[Link] (Z.Z.); 2200180201@[Link] (J.S.)
2 School of Chemical Engineering and Chemistry, Harbin Institute of Technology, Harbin 150001, China;
22b925038@[Link] (Y.W.); wangzhb@[Link] (Z.W.)
* Correspondence: lijunfu@[Link]

Abstract: State of charge (SOC) estimation is an important part of a battery management system (BMS).
As for small portable devices powered by lithium-ion batteries, no current sensor will be configured
in BMS, which presents a challenge to traditional current-based SOC estimation algorithms. In this
work, an electrochemical model is developed for lithium batteries, and three methods, including the
incremental seeking method, dichotomous method, and extended Kalman filter algorithm (EKF),
are separately developed to establish the framework of current and SOC estimation simultaneously.
The results show that the EKF algorithm performs better than the other two methods in terms of
estimation accuracy and convergence speed. In addition, the estimation error of the EKF algorithm is
within ±2%, which demonstrates its feasibility.

Keywords: Li-ion; SOC; current sensorless; extended Kalman filtering

1. Introduction
In recent years, the sustainable development of the environmental ecosystem has
become a pressing issue due to the increasingly prevalent problems of environmental
pollution. In this case, it is essential to focus on the efficient use of energy and resources.
Citation: Zhang, Z.; Shao, J.; Li, J.; Lithium-ion batteries are a popular choice as energy storage components for small electron-
Wang, Y.; Wang, Z. SOC Estimation ics and large electric vehicles due to their high energy density, long cycle life, no memory
Methods for Lithium-Ion Batteries effect, and low self-discharge [1]. However, overcharging or over-discharging a lithium
without Current Monitoring. Batteries battery can lead to capacity degradation, shortened battery life, and even explosion.
2023, 9, 442. [Link] In order to guarantee the safe and secure use of lithium-ion batteries and to extend their
10.3390/batteries9090442 cycle life, a battery management system (BMS) is critical, which can effectively manage the
Academic Editor: Matthieu Dubarry performance of lithium-ion batteries in a comprehensive, efficient, and refined manner [2].
It is necessary to ensure optimal battery performance and longevity in various applications.
Received: 9 August 2023 The reliability of the BMS depends on the precision of the state of charge (SOC)
Revised: 24 August 2023
estimation of lithium-ion batteries. SOC indicates the remaining capacity and provides
Accepted: 28 August 2023
an indication of whether the battery needs to be charged or discharged. The variation
Published: 29 August 2023
characteristics of SOC are a critical performance indicator for assessing the status of lithium-
ion batteries. Therefore, accurate SOC estimation is a core function of BMS [3], and it is a
prerequisite to achieve additional functions such as safety control, battery equalization,
Copyright: © 2023 by the authors.
and troubleshooting.
Licensee MDPI, Basel, Switzerland. However, SOC is an internal state of the battery and cannot be directly measured. It
This article is an open access article can only be estimated based on the relationship between voltage, current, temperature,
distributed under the terms and and the aging of the battery [4]. Therefore, it is crucial to develop reliable SOC estimation
conditions of the Creative Commons algorithms that consider all relevant factors to ensure optimal performance and safety of
Attribution (CC BY) license (https:// lithium-ion batteries in various applications.
[Link]/licenses/by/ In automobiles, robots, and energy storage systems, current measurement is usually
4.0/). achieved using shunt resistors or Hall-effect current sensors. While shunt resistors have

Batteries 2023, 9, 442. [Link] [Link]

Batteries 2023, 9, 442 2 of 20

inherent power losses and require isolation circuitry, Hall-effect sensors are typically
expensive. In low-cost portable applications, such as microphones and hearing aids, the
current sensor is often not equipped, considering the size and cost of the device. Therefore,
it is practical to design a current sensorless SOC estimation method for low-cost portable
applications.
Common SOC estimation algorithms include the open-circuit voltage method, ampere-
hour integral method, electrochemical impedance spectroscopy, adaptive filtering method,
and data-driven neural network method [5]. The open circuit voltage method takes a long
time to collect data and is difficult to apply in practice [6], while the ampere-hour integral
method is affected by the initial charge state and current stability. The electrochemical
impedance spectroscopy method is only used for laboratory research [7]. In addition,
the adaptive filtering method has a complex algorithm and long calculation cycle, which
includes nonlinear Kalman filter, particle filter, specifically including extended Kalman
filter, traceless Kalman filter, and other methods [8–12]. In neural network methods [13–15]
and support vector machine methods [16–19], the SOC estimation of a battery is viewed as
a regression problem, using multiple inputs (e.g., voltage, current, and environmental vari-
ables) to predict the SOC. These methods usually require a large quantity of experimental
data to train the neural network and use various optimization techniques to improve preci-
sion and robustness. Ignoring the internal mechanism of the battery, the model accuracy
depends on the quality of the sample data. In addition, in the field of fast charging, Xuejiao
Xu et al. [20] created their own three-electrode system so as to measure the anode/cathode
potential and estimate the SOC.
A battery model, which includes equivalent circuit models (ECM) [21] and electro-
chemical models, provides high simulation accuracy and reflects the external characteristics
of the battery. The electrochemical model parameters have physical meanings and act as a
bridge between the external characteristics of the battery and the internal electrochemical
reaction mechanism of the battery [22–24]. In order to achieve accurate SOC estimation,
the BMS requires current measurements as input to the estimator.
For the current sensorless SOC estimation method, Cambron and Cramer [25] es-
timated the current by an unknown input observer, and Putra et al. [26] created a new
method to accomplish current estimation on the basis of Thevenin ECM. Chun et al. [27]
obtained the open circuit voltage (OCV) and current information from the terminal voltages
and then calculated the SOC using the ampere-hour integral method. However, these
methods either used a linear relationship between OCV and SOC or utilized an overly
simple battery model that led to a decrease in the accuracy of the model.
To avoid the above problems, Jing Hou et al. [28] used the variational Bayesian
extended Kalman filter method to achieve simultaneous estimation of SOC and current.
Experimental results showed that the mean absolute errors (MAEs) and the root mean
square errors (RMSEs) of the SOC estimations of the proposed variational Bayes-based
unscented Kalman filter (VB-UKF) were less than ±3%.
In this work, three methods are proposed to estimate SOC in the absence of current
sensors based on the electrochemical model of lithium-ion batteries, which include the
incremental seeking method, the dichotomous method, and the improved extended Kalman
filter algorithm. The remainder of this work is organized as follows: (1) In Section 2, the
electrochemical model of lithium-ion batteries is established to mathematically express
the mechanisms in the charging and discharging process, and the relationship function
between voltage and SOC is obtained. The principles of the three methods are described;
(2) In Section 3, the data test of LCO lithium-ion batteries is conducted to obtain the voltage
and current data under different working conditions; (3) In Section 4, the estimation of SOC
under no current monitoring by three methods is completed compared with the reference
value. In addition, the effects of inaccurate initial values of SOC and different levels of
voltage noise on the accuracy of the estimation results are also explored.
 different levels of voltage noise on the accuracy of the estimation results are also explo

2. Lithium-Ion Battery Model and Algorithm Principle

Batteries 2023, 9, 442 2.1. Simplified Electrochemical Model 3 of 20

The structure of a lithium-ion battery is divided into three areas: the positive
trode, the negative electrode, and the diaphragm. The cathode material typically uses
[Link]
Lithium-Ion Battery Model and Algorithm Principle
phosphate (LFP), lithium cobalt oxide (LCO), lithium-nickel-cobalt-ma
2.1. Simplified Electrochemical Model
nese oxide (LNCM), etc., while the anode material is usually graphite, and the electro
The structure of a lithium-ion battery is divided into three areas: the positive electrode,
is mostly LiPF6 alkyl carbonate with polymer materials. During the process of char
the negative electrode, and the diaphragm. The cathode material typically uses lithium
andferrous
discharging,
phosphate lithium
(LFP), ions
lithiumarecobalt
de-embedded
oxide (LCO), and transferred between the positive
lithium-nickel-cobalt-manganese
negative electrodes
oxide (LNCM), etc.,[29].
while the anode material is usually graphite, and the electrolyte is
mostly LiPF6 alkyl
The classical carbonate with polymer
electrochemical model materials. Duringbatteries
of lithium-ion the processis of charging
based on ten con
and discharging, lithium ions are de-embedded and transferred between the positive and
equations in the form of partial differential equations. These equations describe the so
negative electrodes [29].
phase diffusion within
The classical the particles,
electrochemical model liquid-phase
of lithium-iondiffusion
batteries isin the on
based electrolyte,
ten controlsolid-p
potential equilibrium,
equations in the form ofand liquid-phase
partial potentialThese
differential equations. equilibrium
equationsin the positive,
describe the solid-diaphra
phase diffusion within the particles, liquid-phase diffusion
and negative regions. The model structure is shown in Figure 1. in the electrolyte, solid-phase
potential equilibrium, and liquid-phase potential equilibrium in the positive, diaphragm,
and negative regions. The model structure is shown in Figure 1.

Figure 1. Lithium battery electrochemical model structure.

Figure 1. Lithium battery electrochemical model structure.
In this work, the internal processes of the cell were described using algebraic equations.
These equations encompassed various aspects of the cell’s functioning, such as the basic
In this work, the internal processes of the cell were described using algebraic e
working process, solid-phase diffusion process, concentration polarization effect, reaction
tions. These equations
polarization effect, ohmicencompassed various
polarization effect, and theaspects of of
calculation thethecell’s functioning,
terminal voltage [30].such as
basic working
The equations process,
involvedsolid-phase
in the modeldiffusion process,
are presented concentration
in Table polarization
1. The meanings of the effec
parameters involved in the above model are shown in Table 2.
action polarization effect, ohmic polarization effect, and the calculation of the term
voltage [30]. The equations involved in the model are presented in Table 1. The mean
2.2. Identification of Model Parameters
of the parameters involved in the above model are shown in Table 2.
The intrinsic characteristics of the battery parameters remain constant and can be
obtained by consulting the manufacturer or references. However, the variables to be
Table 1. Electrochemical
identified model equations
in the electrochemical model for lithium
include the batteries.
initial lithium intercalation in the
cathode, cathode capacity, offset of the lithium intercalation, and more [31]. The parameters
Correlation Mechanism involved can be obtained by analyzing the intrinsic Equationsconnection between the cell terminal
Terminal voltage Uapp (k)voltage,
voltage change, the open-circuit and- −
= Eocv (k) eachηcon (k)of
part - ηthe (k) - ηohm (k) under the
overpotential
act
designed identification conditions with different forms of current dE excitation.
ref ocv
Open circuit potential correction The parameter identificationEmethod
ocv (k) =referred
Eocv (k) to
+ an
(T(k) - Tref ) response analysis devel-
excitation
oped in [30]. Based on thetksimplified electrochemical model, using the dTinput data of current
2
x (k) =
and the corresponding
avg x - I dt/Q
0 ∫tresponse , x
ndatasurf
(k) = x
of voltage, (k) - ∆x(k), ∆x(k)
avgthe relationship between = ∆x1the
(k)+internal
τ I(k)/Qn ,
1 7 n
mechanisms and external t behaviors was established quantitatively, and the parameters 2 in
Basic working principle y (k) = y0 + ∫t k I dt/Qp , ysurf (k) = yavg (k) + ∆y(k), ∆y(k) = ∆y1 (k) + τp I(k)/Qp ,
the modelavgwere fitted during
1 this process. 7
The parameters obtained byEidentification
(k) = U [y are shown
(k)] - in
UTable
[x 3.(k)]
ocv p surf n surf
1 12τn I(k) 1 12τp I(k)
Solid diffusion ∆x1 (k+1) = ∆x1 (k)+ [ - ∆x1 (k)]∆t, ∆y1 (k + 1) = ∆y1 (k)+ [ - ∆y1 (k)
τn 7Qn τp 7Qp
1 1
Liquid-phase diffusion ∆𝑐n (k+1) = ∆𝑐n (k) + [Pconn I(k) - ∆𝑐n (k)]∆t, ∆𝑐p (k + 1) = ∆𝑐p (k) + [Pconp - ∆𝑐p (k)]∆
τe τe
Batteries 2023, 9, 442 4 of 20

Table 1. Electrochemical model equations for lithium batteries.

Correlation Mechanism Equations

Terminal voltage Uapp (k) = Eocv (k) − ηcon (k) − ηact (k) − ηohm (k)
ref
Open circuit potential correction Eocv (k) = Eocv (k) + ( T (k) − Tref ) dEocv
dT
R tk
xavg (k) = x0 − t1 Idt/Qn , xsurf (k) = xavg (k) − ∆x (k), ∆x (k) = ∆x1 (k) + 72 τn I (k)/Qn ,
Rt
Basic working principle yavg (k) = y0 + t1k Idt/Q p , ysurf (k) = yavg (k) + ∆y(k), ∆y(k) = ∆y1 (k) + 27 τp I (k)/Q p ,
h i h i
Eocv (k) = U p ysurf (k) − Un xsurf (k)
1 12τn I (k ) 12τp I (k)
Solid diffusion ∆x1 (k + 1) = ∆x1 (k) + τn [ 7Qn − ∆x1 (k)]∆t, ∆y1 (k + 1) = ∆y1 (k) + τ1p [ 7Q − ∆y1 (k)]∆t
p
1
∆cn (k + 1) = ∆cn (k) + τe [ Pconn I (k) − ∆cn (k)]∆t, ∆c p (k + 1) = ∆c p (k) + τe Pconp − ∆c p (k) ∆t, 1
 
Liquid-phase diffusion 2RT(k )
h
c + ∆c (k )
i
ηcon (k) = F (1 − t+ )ln c0 − ∆c p (k)
q 0 qn
2RT(k )
ηact (k) = F [ln ( m2n (k) + 1 + mn (k)) − ln( m2p (k) + 1 + m p (k))]
Reactive polarization
m p (k) = 6Q 1c0.5 1
0.5
1
0.5 Pact I ( k ),m n = 6Q c0.5
1
0.5 0.5 Pact I ( k )
p 0 (1 − ysurf ( k )) ( ysurf ( k )) n 0 (1 − xsurf ( k )) ( xsurf ( k ))

Ohmic polarization ηohm = Rohm I (k)

Batteries 2023, 9, 442 5 of 20

Table 2. Meaning of related parameters.

Parameters Physical Meaning

Uapp Terminal voltage (V)
Eocv Electromotive force (V)
ηcon Concentration polarization overpotential (V)
ηact Reaction polarization overpotential (V)
ηohm Ohmic polarization overpotential (V)
Solid-phase average stoichiometric number of positive and negative
yavg , xavg
electrodes (-)
Solid-phase surface stoichiometric number of positive and negative
ysurf , xsurf
electrodes (-)
∆y Deviations between ysurf and yavg (-)
∆x Deviations between xsurf and xavg (-)
∆y1 Intermediate variable of ∆y (-)
∆x1 Intermediate variable of ∆x (-)
T Battery internal temperature (K)
R Ideal gas constant (J mol−1 K−1 )
Initial lithium intercalation concentration fraction of positive and
y0 , x0
negative electrodes (-)
c0 Initial lithium ion concentration in electrolyte (mol m−3 )
Qp , Qn Total capacity of positive and negative electrodes (A s)
Qall Total capacity of the battery (A s)
τp , τn Solid diffusion time constants of positive and negative electrodes (s)
Dx , Dy Lithium embedding rate of positive and negative electrodes (-)
Pact Reaction polarization coefficient (m−1.5 mol0.5 s)
Pcon Proportional coefficient of liquid phase diffusion (mol m−3 A−1 )
Rohm Ohm internal resistance (Ω)
yofs Embedded lithium offset (-)
τe Liquid phase diffusion time constant (s)

Table 3. Battery parameters.

Parameters Value
y0 , x0 0.7941, 0.4538
c0 1000
Qp , Qn 19,859.04, 11,888.64
τp , τn 184.7533, 2.5501
Dx , Dy 0.7922, 0.4743
Pact 271,780
Pcon 955.1863
Rohm 0.0686
yofs 0.0708
τe 62.2686

2.3. Principle of SOC Estimation Algorithm under No Current Monitoring

The three SOC estimation algorithms utilized in this study were based on the simplified
electrochemical model. However, the model function inputs did not take into account the
temperature of the battery. The inputs solely consisted of the SOC and current, while the
output was limited to the cell terminal voltage.

2.3.1. Incremental Seeking Method

This solution was based on the idea of enumeration to achieve current estimation
and SOC prediction. The approach involves gradually increasing the current within fixed
boundary conditions and then inputting the current into the model to calculate the voltage
to identify the current values that meet the necessary requirements. Once the appropriate
current value is determined, the SOC can be estimated using the ampere-hour integral
method. The corresponding flowchart illustrating this method is depicted in Figure 2.
Batteries 2023, 9, x FOR PEER REVIEW 6 of 20
Batteries 2023, 9, 442 6 of 20

Figure 2. Flow chart of incremental seeking method.

Figure 2. Flow chart of incremental seeking method.
The specific scheme was as follows: Firstly, the measured voltage was compared with
theThe
previous
specific second’s
schemevoltage
was astofollows:
determine whether
Firstly, the batteryvoltage
the measured was being
wascharged
comparedor with
thedischarged. Then, the current sign was initialized, and upper and lower search boundaries
previous second’s voltage to determine whether the battery was being charged or dis-
were set. An initial current value of 0 was assumed. A voltage simulation was carried
charged. Then, the current sign was initialized, and upper and lower search boundaries
out using the assumed current, and the simulated voltage was compared to the measured
were set. An
voltage. initial
If the current
error value
between the of 0 was assumed.
measured voltage andA voltage simulation
simulated was carried
voltage exceeded an out
using the assumed current, and the simulated voltage was compared to the
acceptable range, the current value was increased based on the error until the error was measured
voltage.
within Ifanthe error between
acceptable the measured
range. The voltage andtosimulated
current corresponding voltage
the simulated exceeded
voltage was an
acceptable
considered range, the current
the current, value
and the SOCwaswasincreased
predicted based on ampere-hour
using the the error until the error was
integration
method.
within an acceptable range. The current corresponding to the simulated voltage was con-
sidered the current, and the SOC was predicted using the ampere-hour integration
2.3.2. Dichotomous Method
method.
As shown in Figure 3, this scheme used the dichotomy method for current estimation
and SOC prediction. The basic principle of the dichotomy method involved continuously
2.3.2. Dichotomous Method
dividing the search range into two parts within fixed boundary conditions to find the
As shown
number in Figure
that meets 3, requirements.
certain this scheme used the dichotomy method for current estimation
and SOC prediction. The basic principle of the dichotomy method involved continuously
dividing the search range into two parts within fixed boundary conditions to find the
number that meets certain requirements.
Batteries 2023, 9, x FOR PEER REVIEW
Batteries 2023, 9, 442 7 of 20

Figure 3. Dichotomous search flow chart.

Figure 3. Dichotomous search flow chart.
The specific scheme is as follows: firstly, the measured voltage was compared with that
of the previous second to determine the charging and discharging state at this time. Then,
the The specific
current schemeand
was initialized, is the
as follows:
upper andfirstly, the measured
lower search boundaries werevoltage was The
assumed. compared
that of the
voltage previous
simulation wassecond
carried to
outdetermine
based on thethe charging
assumed and
current, anddischarging state at this
the corresponding
simulation voltage was compared with the measured voltage. If
Then, the current was initialized, and the upper and lower search boundaries the error is greater than we
the acceptable range, the current value is adjusted according to the
sumed. The voltage simulation was carried out based on the assumed current, an error until the error
between the measured voltage and the simulated voltage is in the acceptable range. At this
corresponding simulation voltage was compared with the measured voltage. If the
point, the current corresponding to the simulated voltage was considered as the current at
is that
greater thanFinally,
moment. the acceptable range, was
the SOC prediction thecarried
currentoutvalue
by theis adjusted according
ampere-hour integration to the
until the error between the measured voltage and the simulated voltage is in the acc
method.
ble range. At this point, the current corresponding to the simulated voltage was co
2.3.3. Extended Kalman Filter
ered as the current at that moment. Finally, the SOC prediction was carried out b
The solution utilized a modified extended Kalman filter to achieve SOC estimation.
ampere-hour integration
Based on the simplified method. model, the load current was taken as the unknown
electrochemical
input, and the system state equation and the measurement equation were established. In
2.3.3. Extended
addition, Kalman
the current Filterwas achieved using the modified extended Kalman filter
estimation
method. Finally, the SOC estimation was completed by the ampere-hour integral method.
The solution utilized a modified extended Kalman filter to achieve SOC
The flow chart is shown in Figure 4.
estim
Based on the simplified electrochemical model, the load current was taken as th
known input, and the system state equation and the measurement equation were
lished. In addition, the current estimation was achieved using the modified extended
man filter method. Finally, the SOC estimation was completed by the ampere-hour
gral method. The flow chart is shown in Figure 4.
Batteries 2023,
Batteries 2023, 9,
9, 442
x FOR PEER REVIEW 88 of
of 20
20

Figure 4. Extended Kalman filter estimation flow chart.

To implement
implementthis thissolution,
solution,ananupdate
updateof of state
state variables
variables andand observation
observation updates
updates was
was required, which meant that the priori SOC estimation, succession of
required, which meant that the priori SOC estimation, succession of current values, current values,
and
and covariance
covariance matrix
matrix calculations
calculations werewere carried
carried out. out. The gain
The gain matrixmatrix was then
was then calculated,
calculated, and
and the currents
the currents were were corrected
corrected according
according to voltage
to voltage [Link].
Finally,Finally, the accurate
the accurate SOC
SOC estima-
estimation for a single
tion for a single cell
cell was was obtained
obtained by the by the ampere-hour
ampere-hour integration
integration [Link].

3. Battery Testing Process and Results

3. Battery Testing Process and Results
Different lithium cobaltate battery monoblocks from the same batch with a standard
Different lithium cobaltate battery monoblocks from the same batch with a standard
capacity of 800 mAh were selected. Charging and discharging tests were performed using
capacity of 800 mAh were selected. Charging and discharging tests were performed using
Neware’s battery test system to simulate the battery operating conditions at different
Neware’s battery test system to simulate the battery operating conditions at different
charge and discharge multipliers. The data sampling frequency was 1 s. The specific
charge and discharge multipliers. The data sampling frequency was 1 s. The specific cur-
current configurations were as follows:
rent configurations were as follows:
The specific current configuration for the DST operating conditions test is described as
The specific current configuration for the DST operating conditions test is described
follows:
as follows:
(1) The lithium-ion battery was fully charged by constant current and constant voltage
(1) The lithium-ion battery was fully charged by constant current and constant voltage
charging.
charging.
(2) The battery was rested in a constant temperature chamber for 1 h.
(2) The
(3) The battery
battery was
was rested in a constant
discharged temperature
at a constant chamber
discharge for [Link] 30 s, discharged
rate of 0.25
(3) The battery was discharged at a constant discharge
at 0.5 C for 12 s, and charged at 0.25 C for 10 s. rate of 0.25 C for 30 s, discharged
(4) at 0.5(3)
Step C for
was12repeated
s, and charged at 0.25 C for 10 s.
three times.
(4) Step (3) was repeated three times.
(5) The battery was charged successively at a constant charge rate of 0.25 C for 35 s, 2 C
(5) The battery
for 10 s, andwas
1.25charged
C for 25successively
s. at a constant charge rate of 0.25 C for 35 s, 2 C
for 10 s, and 1.25 C for 25 s.
(6) The battery was charged at a constant charge rate of 0.5 C for 10 s, discharged at 0.5 C
(6) The battery
for 30 s, andwas charged
charged at 1 at
C afor
constant
10 s. charge rate of 0.5 C for 10 s, discharged at 0.5
(7) C forbattery
The 30 s, and
wascharged at 150
rested for C s.
for 10 s.
(7) The battery was rested for 50 s.
 (8) Steps (3)–(7) were repeated until the voltage decreased to 3.7 V. Then, the batte
discharged at a constant discharge rate of 0.5 C until the voltage was 2.5 V to s
experiment.

Batteries 2023, 9, 442

The HPPC condition test with the specific current configuration 9was
of 20
condu
follows:
(1) The lithium-ion battery was fully charged by constant current and constant
(8) charging.
Steps (3)–(7) were repeated until the voltage decreased to 3.7 V. Then, the battery was
discharged
(2) The batteryat awas
constant discharge
rested rate of 0.5temperature
in a constant C until the voltage was 2.5for
chamber V to1stop
h. the
experiment.
(3) The battery was discharged at 1 C for 8 min and was rested for 30 min.
The HPPC condition test with the specific current configuration was conducted as
(4) The battery was discharged at 3 C for 10 s and rested for 3 min.
follows:
(5) The battery was charged at 1.5 C for 10 s and rested for 2 min.
(1) The lithium-ion battery was fully charged by constant current and constant voltage
(6) Steps (3)–(5) were repeated once.
charging.
(7)
(2) The current
The battery wasinrested
(4) and
in a (5) was temperature
constant modified to 1.875 C.
chamber for Steps
1 h. (3)–(5) were repeate
(3) the
The battery was discharged at 1 C
resting voltage dropped to 3.6 V. for 8 min and was rested for 30 min.
(4) The battery was discharged at 3 C for 10 s and rested for 3 min.
(8) The battery was discharged at 0.9 C until the voltage was 2.5 V to stop the expe
(5) The battery was charged at 1.5 C for 10 s and rested for 2 min.
(6) Steps (3)–(5) were repeated once.
4.
(7)Results and Discussion
The current in (4) and (5) was modified to 1.875 C. Steps (3)–(5) were repeated until
the resting voltage
It is worth noting dropped to three
that all 3.6 V. methods of SOC estimation without current m
(8) The battery was discharged at 0.9 C until the voltage was 2.5 V to stop the experiment.
ing include the ampere-hour integration method, and the estimation accuracy is
influenced
4. Results andbyDiscussion
the initial SOC accuracy and noise. The effects of different initial SO
random voltage
It is worth noise
noting thaton the algorithm
all three methods of estimation results
SOC estimation were
without shown
current and discu
monitor-
follows,
ing includerespectively.
the ampere-hour integration method, and the estimation accuracy is greatly
influenced by the initial SOC accuracy and noise. The effects of different initial SOCs and
random voltage noise on the algorithm estimation results were shown and discussed as
4.1. Effect of Different Initial SOC on Estimation Results
follows, respectively.
The initial SOC of the test data was set as 100%, 70%, and 90%, respectively. T
4.1. Effect of Different Initial SOC on Estimation Results
mated results were as follows.
The initial SOC of the test data was set as 100%, 70%, and 90%, respectively. The
estimated results were as follows.
4.1.1. Incremental Seeking Method
4.1.1. The
Incremental Seeking
estimation Method
results using the incremental seeking method for different init
statesThe
areestimation
shown in results using5the
Figures incremental
and 6. seeking method for different initial SOC
states are shown in Figures 5 and 6.

Figure 5.5. Estimation

Figure Estimation results of incremental
results seeking
of incremental methodmethod
seeking under DST conditions
under and differentand diffe
DST conditions
initial SOC.
itial SOC.
Batteries 2023, 9, x FOR PEER REVIEW 10 of 20

x FOR PEER REVIEW

Batteries 2023, 9, 442 10 of 20

Figure 6. Estimation results of incremental seeking method under HPPC conditions and different
initial SOC.
Figure 6. Estimation results of incremental seeking method under HPPC conditions and different
initial SOC.
initialThe estimated trend obtained by the incremental seeking method was consistent with
SOC.
the actual reference value, and the estimated results varied with the battery charge and
The estimated trend obtained by the incremental seeking method was consistent with
discharge state. For example, under DST conditions, SOC decreased along a wave line
the actual reference value, and the estimated results varied with the battery charge and
until 12,500 s and then decreased along a straight line. Under HPPC conditions, SOC de-
discharge state.
state. For
Forexample,
example,under
underDSTDSTconditions,
conditions,SOCSOCdecreased
decreasedalong a wave
along line until
a wave line
creased
12,500
along a
s and sthen
step.
until 12,500 anddecreased along along
then decreased a straight line. Under
a straight HPPCHPPC
line. Under conditions, SOC decreased
conditions, SOC de-
alongHowever,
creased a step. the difference between the estimation results with initial SOC values of
along a step.
100%,However,
90%, and
However, the 70% was significant
difference between under
between the HPPC
estimation
the estimation condition.
results The SOC
with initial errorvalues
was largest
of
when the initial SOC was accurate, illustrating
100%, 90%, and 70% was significant under the HPPC the instability
HPPC condition. of
condition. Thethe algorithm. When the
The error was largest
initial
when the SOC stateSOC
initial waswas inaccurate,
accurate, the incremental
illustrating seeking of
the instability method converged
the algorithm. Whenwhen the the
SOC
initial SOC state was inaccurate, the incremental seeking method converged when theby a
dropped
initial SOC to
state nearly
was 80%,
inaccurate, and
the the convergence
incremental process
seeking method was often
converged accompanied
when the SOC
large
dropped
SOC abrupt
dropped change.
to nearly 80%,
to nearly After
andconvergence
80%, the
andconvergencewas process
the convergencecompleted,
was the
process initial
often
was SOC
accompanied
often valuebyhad
accompanied a large
a large
by a
abruptabrupt
impact
large change. After convergence
on thechange.
estimationAfterresults. was completed,
convergence the initial
was completed, the SOC
initialvalue
SOChad a large
value had aimpact
large
on the estimation
impact results. results.
on the estimation
4.1.2. Dichotomous Method
4.1.2. Dichotomous Method
[Link]
Dichotomous
estimationMethodresults using the dichotomous method for different initial SOC states
The estimation results using the dichotomous method for different initial SOC states
The estimation
are shown in Figures results
7 andusing
8. the dichotomous method for different initial SOC states
are shown in Figures 7 and 8.
are shown in Figures 7 and 8.

Figure 7.
Figure
Figure 7. Estimation results
7. Estimation
Estimation results of
resultsof dichotomous
ofdichotomous method
dichotomousmethod
methodunder
under
under DST
DST
DST conditions andand
conditions
conditions and different initial
different
different SOC.
initial
initial SOC. SOC.
Batteries 2023,
Batteries 2023, 9,
9, 442
x FOR PEER REVIEW 11 of
11 of 20
20

Figure 8. Estimation results of dichotomous method under HPPC conditions and different initial
SOC.

When applying the dichotomy method, the overall estimation trend was consistent
Figure 8.
8. Estimationresults
resultsofof dichotomous method under HPPC conditions and different
initialinitial
theEstimation
Figure
with actual reference dichotomous
value, and method under HPPC
the algorithmic conditions
estimation and different
results captured SOC.
the same
SOC.
change phenomenon as the charging and discharging state
When applying the dichotomy method, the overall estimation trend was consistent of the battery varied.
with From
When the
the actual overall
applying theviewpoint,
reference value, and the
dichotomy the dichotomous
algorithmic
method, SOC
estimation
the overall estimation
estimationresults error
captured
trend was large, espe-
the same
was consistent
cially
change under HPPC
phenomenon conditions,
as the and
charging andsignificant
discharging estimation
state of thebias could
battery
with the actual reference value, and the algorithmic estimation results captured the same be
[Link] in Figure 8.
When
changeFrom
the the overall
initial
phenomenon SOC viewpoint,
as state wasthe
the charging dichotomous
inaccurate, SOC
if the
and discharging estimation
initial
state ofSOC error90%,
was
the battery was large,
it
varied. espe- con-
completed
ciallyFrom
vergenceunder
whenHPPC
the theconditions,
overall viewpoint,
SOC and
decreasedthesignificant
to aroundestimation
dichotomous bias could
SOCInestimation
85%. addition, ifbe
errorthe seen
was in Figure
large,
initial SOC 8. 70%,
espe-
was
When
cially the initial
under SOC
HPPC state was
conditions, inaccurate,
and if
significant the initial SOC
estimation was
bias
it completed convergence when the SOC decreased to around 75%. The convergence pro- 90%,
could it
be completed
seen in conver-
Figure 8.
gence
When when the SOC decreased
theainitial to around 85%. Ininitial
addition, ifwas
the initial itSOC was 70%, it
cess had greatSOC stateof
degree was inaccurate,
abrupt change, if the
but it wasSOC less 90%,the
than completed
incremental con- seeking
completed convergence when the SOC decreased to around 75%. The
vergence when the SOC decreased to around 85%. In addition, if the initial SOC was 70%, convergence process
method.
had Inaccurate
a great degree initial
of abrupt SOC had
change, a small impact
but itdecreased
was less than on the estimation
the incremental results after
seeking method. the con-
it completed convergence when the SOC to around 75%. The convergence pro-
vergence
Inaccurateof the algorithm.
initial SOC had a small impact on the estimation results after the convergence of
cess had a great degree of abrupt change, but it was less than the incremental seeking
the algorithm.
method. Inaccurate initial SOC had a small impact on the estimation results after the con-
4.1.3. EKFof the algorithm.
vergence
4.1.3. EKF
The estimation results using the EKF algorithm for different initial SOC states are
[Link]
shown inestimation
EKF Figures 9 results
and [Link] the EKF algorithm for different initial SOC states are
shown in Figures 9 and 10.
The estimation results using the EKF algorithm for different initial SOC states are
shown in Figures 9 and 10.

Figure 9.
Figure Estimation
Estimationresults of of
EKF algorithm for DST conditions and different initial SOC.
Figure 9.
9. Estimation results
results of EKFEKF algorithm
algorithm for DST
for DST conditions
conditions and different
and different initial SOC.
initial SOC.
 Batteries2023,
Batteries 2023,9,9,442
x FOR PEER REVIEW 12of
12 of20
20

10. Estimation
Figure10.
Figure Estimation results
resultsof
ofEKF
EKFalgorithm
algorithmfor
forHPPC
HPPCconditions
conditionsand
anddifferent
differentinitial
initialSOC.
SOC.

The
The overall
overall estimation
estimation trend
trend obtained
obtained using
using the
the EKF
EKF algorithm
algorithm was
was consistent
consistent with
with
the actual reference value, and the algorithm estimation results
the actual reference value, and the algorithm estimation results reflected thereflected the same variation
same varia-
phenomenon
tion phenomenonas theasbattery charge
the battery and discharge
charge state change.
and discharge state change.
As a whole, the SOC estimation error
As a whole, the SOC estimation error of the EKFof the EKF algorithm
algorithm was small,
was andand
small, thethe
average
aver-
errors for the
age errors fortwo
the operating conditions
two operating whenwhen
conditions the initial SOC SOC
the initial was accurate werewere
was accurate 1.94%1.94%
and
1.13%, respectively. When the initial SOC was inaccurate, the algorithm
and 1.13%, respectively. When the initial SOC was inaccurate, the algorithm converged converged quickly
except
quicklyfor the DST
except for condition where thewhere
the DST condition initialthe
SOC is 70%.
initial SOC Convergence was achieved
is 70%. Convergence was
when
achieved when the SOC decreased to about 85~90%. The convergence processsmooth,
the SOC decreased to about 85~90%. The convergence process was and
was smooth,
the
andconvergence
the convergence estimates werewere
estimates largely consistent
largely withwith
consistent the reference values
the reference and almost
values and al-
independent of the initial SOC.
most independent of the initial SOC.
The estimation errors of the three methods for different initial SOC states are shown in
The estimation errors of the three methods for different initial SOC states are shown
Table 4.
in Table 4.
Table 4. Estimation error at different initial SOC states.
Table 4. Estimation error at different initial SOC states.
Accurate Initial SOC
Accurate Initial SOC InitialInitial
SOC 0.9SOC 0.9 Initial SOC
Initial 0.7 0.7
SOC
Incremental seeking
Incremental method
seeking method soc soc VoltageVoltage soc soc Voltage
Voltage soc soc Voltage
Voltage
Average value 3.58% 0.042 V 3.60% 0.035 V 5.08% 0.043 V
DST Average value 3.58% 0.042 V 3.60% 0.035 V 5.08% 0.043 V
DST Maximum value 12.76% 0.300 V 11.42% 0.450 V 30% 0.659 V
Maximum value
Average value 7.62%12.76% 0.025 V 0.300 V 7.80% 11.42% 0.028 0.450
V V 30%
9.10% 0.659VV
0.035
HPPC Average value
Maximum value 15.13%7.62% 0.373 V 0.025 V15.20% 7.80% 0.028
0.372 V V 9.10%
30% 0.035VV
0.669
HPPC
Maximum
Dichotomous method value soc 15.13% Voltage 0.373 V soc 15.20% Voltage 0.372 V soc30% 0.669 V
Voltage
Dichotomous method
Average value 1.87% soc 0.014 VVoltage 2.58% soc 0.016Voltage
V 4.17%soc Voltage
0.084 V
DST
Maximum
Averagevalue
value 5.02% 1.87% 0.416 V 0.014 V10.00% 2.58% 0.416 0.016
V V 30%
4.17% 0.495 V
0.084 V
DST Average value 4.51% 0.028 V
HPPC Maximum value 5.02% 0.416 V 4.43% 10.00% 0.030 0.416
V
V 6.37%
30% 0.044 V
0.495 V
Maximum value 8.82% 0.352 V 10.00% 0.296 V 30% 0.288 V
Average value 4.51% 0.028 V 4.43% 0.030 V 6.37% 0.044 V
HPPC EKF soc Voltage
Maximum value 8.82% 0.352 V soc 10.00% Voltage0.296 V soc
30% Voltage
0.288 V
Average value 1.94% 0.034 V 2.36% 0.035 V 3.82% 0.036 V
DST EKF
Maximum value 5.27% soc 0.414 V Voltage10.21% soc Voltage
0.414 V 30.15%soc Voltage
0.414 V
Average
Average value
value 1.13% 1.94% 0.0098 V0.034 V 1.41% 2.36% 0.010 0.035
V V 2.33%3.82% 0.036VV
0.011
DST
HPPC
Maximum value 3.69% 5.27% 0.3872 V0.414 V10.20% 10.21% 0.387 0.414
V
Maximum value V 30.14%
30.15% 0.387
0.414VV
Average value 1.13% 0.0098 V 1.41% 0.010 V 2.33% 0.011 V
HPPC
Maximum valueIn terms of3.69%
overall SOC0.3872 V performance,
estimation 10.20% 0.387
the EKFV algorithm
30.14%performed
0.387 Vthe
best and met the actual demand. The dichotomy method was the second-best and could
In terms
accurately of overall
estimate SOCSOC
whenestimation
the initialperformance, the EKF
state was reliable. algorithmthe
However, performed the
incremental
best andmethod
seeking met thehad
actual
thedemand. The dichotomy
worst performance, withmethod was the
the highest second-best
error, failing toand could
meet the
accurately
actual estimate
demand SOC when
and requiring the initial
further state was reliable. However, the incremental
improvement.
 algorithm was highly resistant to interference and almost unaffected. The dich
method was less affected, while the incremental seeking method was the most un
Inaccurate initial states had a significant impact on the results of the incremental s
Batteries 2023, 9, 442 method. 13 of 20
Concerning the convergence of algorithms under different initial SOC states, t
converged fast and seamlessly, while the dichotomy method converged slightly
thanRegarding
the EKF. theIn contrast, the incremental
impact of different initial SOCseeking method had
states on algorithm the slowest
estimation, the EKFconvergen
al-
gorithm was highly resistant to interference and almost unaffected.
with a sudden change in the convergence process caused by the algorithm’s princip The dichotomy method
was less affected, while the incremental seeking method was the most unstable. Inaccurate
If the initial SOC error was too large, there was a significant increase in the e
initial states had a significant impact on the results of the incremental seeking method.
all three methods.
Concerning This indicates
the convergence that theunder
of algorithms errordifferent
correction
initialcapability of the
SOC states, the EKFthree m
isconverged
limited due to the
fast and lack of while
seamlessly, current the input. However,
dichotomy method it is worthslightly
converged notingslower
that the initi
than the EKF. In contrast, the incremental seeking method had
error is usually not so large in portable devices. Furthermore, the SOC error canthe slowest convergence
rate, with
brated a sudden change
gradually duringinthe the convergence
rest [Link] caused by the algorithm’s principle.
If the initial SOC error was too large, there was a significant increase in the error for
Jingmethods.
all three Hou [24] Thisetindicates
al. conductedthat thepulse discharge
error correction experiments
capability to validate
of the three methodstheir pr
VB-UKF
is limited method.
due to the The lack average
of currentabsolute errors itofisthe
input. However, twonoting
worth methods, VB-UKF
that the initial and
scented
SOC errorrecursive
is usually notthree-step
so large in filter (URTSF),
portable [Link] 1.52% and
Furthermore, 2.65%
the SOC when
error can bethe initi
calibrated gradually during the rest mode.
was accurate. When the initial SOC was 80%, the errors of the two methods were
Jing Hou et al. [24] conducted pulse discharge experiments to validate their proposed
and 2.28%, respectively. When the initial SOC was 60%, the errors of the two m
VB-UKF method. The average absolute errors of the two methods, VB-UKF and the
were 2.16%
unscented and 2.58%,
recursive three-steprespectively.
filter (URTSF), Itwere
can 1.52%
be seen
and that
2.65%thewhen estimation effect of t
the initial SOC
algorithm
was accurate. is When
perfectly acceptable.
the initial SOC was 80%, the errors of the two methods were 1.67%
and 2.28%, respectively. When the initial SOC was 60%, the errors of the two methods were
2.16%
4.2. and 2.58%,
Effect respectively.
of Different VoltageIt can be seen
Noise that the estimation
on Algorithm effect Results
Estimation of the EKF algorithm
is perfectly acceptable.
The test data was free of voltage noise, and random Gaussian noise with the m
04.2.
andEffect
3σofofDifferent
10 mV, Voltage Noiseand
50 mV, on Algorithm
100 mVEstimation Results
was superimposed, respectively. The est
The test data was
results are shown below. free of voltage noise, and random Gaussian noise with the mean
of 0 and 3σ of 10 mV, 50 mV, and 100 mV was superimposed, respectively. The estimated
results are shown below.
4.2.1. Incremental Seeking Method
4.2.1. The
Incremental Seeking Method
SOC estimation results under different voltage noise using the increment
The SOC estimation
ing method are shown results under different
in Figures 11 andvoltage
12. noise using the incremental seeking
method are shown in Figures 11 and 12.

[Link]
Figure 11.
Figure results
Estimation of incremental
results seeking method
of incremental seekingunder DST under
method conditions
DSTandconditions
different and d
voltage noise.
voltage noise.
Batteries 2023, 9, x FOR PEER REVIEW
Batteries 2023, 9, 442 14 of 20
14 of 20

Figure 12. Estimation results of incremental seeking method under HPPC conditions and d
voltage noise.

Figure
FigureWhen
12. applying
[Link]
Estimation the
results
results incremental
of incremental
of incremental seeking
seeking method
method
seeking under
under under
method HPPC voltage
conditions
HPPC noise,
and the
and different
conditions esti
differen
voltage
voltage noise.
noise.
trend was consistent with the actual reference value, and the estimation results re
the same
Whenchanges as the
applying the batteryseeking
incremental charging andunder
method discharging states.
voltage noise, the estimation
When applying the incremental seeking method under voltage noise, the estimation
trendHowever, thewith
was consistent presence of voltage
the actual referencenoise
value, caused more irregular
and the estimation fluctuations in
results reflected
trend
the
was
same
consistent
changes as the
with thecharging
battery
actual reference
and
value,states.
discharging
and the estimation results reflected
gorithm’s estimation accuracy, resulting in large errors. This suggests that the incre
the same changes
However, as the battery
the presence charging
of voltage noise and discharging
caused [Link] in the
more irregular
seeking method is susceptible to noise.
However, the presence of voltage noise caused more irregular fluctuations in the al-
algorithm’s estimation accuracy, resulting in large errors. This suggests that the incremental
gorithm’s
seeking estimation
method accuracy,
is susceptible resulting in large errors. This suggests that the incrementa
to noise.
4.2.2. Dichotomous
seeking Method to noise.
method is susceptible
4.2.2. Dichotomous Method
The SOC estimation results using the dichotomous method with different
[Link]
noises
SOC estimation results using the dichotomous method with different voltage
Dichotomous
are shown inMethod
Figures 13 and 14.
noises are shown in Figures 13 and 14.
The SOC estimation results using the dichotomous method with different voltage
noises are shown in Figures 13 and 14.

[Link]
Figure 13.
Figure Estimationresults of dichotomous
results method under
of dichotomous DST conditions
method under andDSTdifferent voltageand
conditions [Link]
Figure 13. Estimation results of dichotomous method under DST conditions and different voltage
noise.
noise.
Batteries 2023, 9, x FOR PEER REVIEW 15 of 2
Batteries 2023, 9, 442 15 of 20

Figure 14. Estimation results of dichotomous method under HPPC conditions and different
noise.

Figure
FigureWhen
14. usingresults
[Link]
Estimationthe dichotomous
of dichotomous
results method
method
of dichotomous under
under
method HPPCvoltage
under noise,
conditions
HPPC the
andestimation
and different
conditions tre
voltage voltage
different
consistent
noise.
noise. with the actual reference value, and the estimation result reflected th
changes as the battery charging and discharging states.
When using the dichotomous method under voltage noise, the estimation trend was
When
The usingerror
mean the dichotomous method under voltage noise, the the
estimation trend was
consistent with the actualofreference
SOC estimation is small
value, and the when
estimation using
result reflected dichotomous
the same m
consistent
However, with the
appropriate
changes as the
actual reference
voltage
battery charging
value,
noise could
and discharging
and the estimation result reflected
improve the estimation accuracy of th
states.
the same
changes
rithm, as
Thewhilethe battery
meanexcessive charging
error of SOCnoise and discharging
led to significant
estimation is small when states.
deviations in the results
using the dichotomous under the D
method.
The
However, mean error
appropriate of SOC
voltage estimation
noise could is
improvesmall
the when using
estimation the of
accuracy dichotomous
the algorithm,method
dition. This shows the poor stability of the dichotomous method.
However,
while appropriate
excessive voltage
noise led to noisedeviations
significant could improve the estimation
in the results accuracy
under the DST of the algo
condition.
This shows
rithm, theexcessive
while poor stability
noiseof led
the to
dichotomous
significantmethod.
deviations in the results under the DST con
4.2.3. EKF
dition. This shows the poor stability of the dichotomous method.
4.2.3. EKF
The SOC estimation results using the EKF algorithm with different voltage n
[Link]
shown EKFSOC estimation results using the EKF algorithm with different voltage noise are
in Figures 15 and 16.
shown in Figures 15 and 16.
The SOC estimation results using the EKF algorithm with different voltage noise are
shown in Figures 15 and 16.

[Link]
Figure 15.
Figure Estimationresults of theof
results EKF
thealgorithm for DST conditions
EKF algorithm for DST and different voltage
conditions noise. voltage
and different
Figure 15. Estimation results of the EKF algorithm for DST conditions and different voltage noise.
Batteries 2023,
Batteries 2023, 9,
9, 442
x FOR PEER REVIEW 16 of 20

Figure 16. Estimation results of the EKF algorithm for HPPC conditions and different voltage noise.

When using
usingthetheEKF
EKFmethod
method under
under voltage
voltage noise,
noise, the the estimation
estimation trend
trend remained
remained con-
consistent
sistent with with
thethe actual
actual reference
reference value,
value, andthe
and theestimation
estimationresults
resultscould
could keep
keep the same
change
change when
when the
the charge
charge and
and discharge
discharge state
state of
of the
the battery
battery changed.
changed.
Overall,
Overall, the EKF algorithm demonstrated a high estimation
the EKF algorithm demonstrated a high estimation accuracy
accuracy even
even under
under
voltage noise, and the SOC estimation results were consistently reliable
voltage noise, and the SOC estimation results were consistently reliable despite thedespite the presence
pres-
of different
ence levelslevels
of different of voltage [Link].
of voltage
The
The estimation
estimation errors
errorsof ofthe
thethree
threemethods
methodsunderunderdifferent voltage
different voltagenoise areare
noise shown
shownin
Table 5.
in Table 5.
Table 5. Estimation error under different voltage noise.
Table 5. Estimation error under different voltage noise.
No Noise
No Noise 10 10
mVmV
Noise
Noise 5050mV
mVNoise
Noise 100 mV
100 mVNoise
Noise
Incremental
Incremental seeking
seeking method soc soc Voltage
method Voltage socsoc Voltage Voltage socsoc Voltage
Voltage soc
soc Voltage
Voltage
Average value
Average value 3.58%3.58% 0.042 V
0.042 V 5.92%
5.92% 0.044 V
0.044 V 5.81%
5.81% 0.051 V
0.051 V 5.01%
5.01% 0.060 V
0.060 V
DST
DST Maximum value 12.76% 0.300 V 11.99% 0.422 V 11.31% 0.342 V 11.14% 0.418 V
Maximum value 12.76% 0.300 V 11.99% 0.422 V 11.31% 0.342 V 11.14% 0.418 V
Average value 7.62% 0.025 V 6.63% 0.026 V 6.63% 0.033 V 7.27% 0.056 V
HPPC Averagevalue
value 15.13% 7.62% 0.3730.025 V 14.78%
6.63% 0.411 0.026 V 14.86%
6.63% 0.374 0.033VV 15.58%
7.27% 0.056 V
HPPC Maximum V V 0.406 V
Maximum value 15.13% 0.373 V 14.78% 0.411 V 14.86% 0.374 V 15.58% 0.406 V
Dichotomous method soc Voltage soc Voltage soc Voltage soc Voltage
Dichotomous method
Average value 1.87% soc Voltage
0.014 V 3.66%soc Voltage
0.077 V soc
1.83% Voltage
0.074 V soc
1.56% Voltage
0.082 V
DST
Averagevalue
Maximum value 5.02%1.87% 0.4160.014 V V 7.34% 3.66% 1.020 0.077
V V 5.09% 1.83% 0.685 0.074VV 1.56%
7.19% 0.082 V
0.732 V
DST Average value 4.51% 0.028 V 1.38% 0.024 V 1.54% 0.040 V 1.58% 0.057 V
HPPC Maximum value 5.02% 0.416 V 7.34% 1.020 V 5.09% 0.685 V 7.19% 0.732 V
Maximum value 8.82% 0.352 V 5.74% 0.937 V 7.18% 0.520 V 6.96% 0.539 V
Average value 4.51% 0.028 V 1.38% 0.024 V 1.54% 0.040 V 1.58% 0.057 V
HPPC EKF
Maximum value soc 8.82% Voltage 0.352 V soc 5.74% Voltage
0.937 V soc 7.18% Voltage
0.520 V soc
6.96% Voltage
0.539 V
Average value 1.94% 0.034 V 1.94% 0.034 V 1.96% 0.037 V 1.90% 0.045 V
DST EKF
Maximum value 5.27%
soc Voltage
0.414 V 5.30%
soc Voltage
0.420 V
soc
5.42%
Voltage
0.448 V
soc
4.98%
Voltage
0.415 V
Average
Average value 1.13%1.94%0.0098
value 0.034
V V 1.13% 1.94% 0.011 0.034
V V 1.11% 1.96% 0.020 0.037VV 1.90%
1.11% 0.045 V
0.032 V
DST
HPPC
Maximum
Maximum value 3.69%5.27%0.3872
value 0.414
V V 3.67% 5.30% 0.384 0.420
V V 3.69% 5.42% 0.375 0.448VV 4.98%
3.76% 0.415 V
0.359 V
Average value 1.13% 0.0098 V 1.13% 0.011 V 1.11% 0.020 V 1.11% 0.032 V
HPPC
Maximum value 3.69% 0.3872 V 3.67% 0.384 V 3.69% 0.375 V 3.76% 0.359 V
From the experimental results, we can see that the error of the EKF algorithm under
different noises was within 2%, which would meet the practical requirements. The dichoto-
mousFrommethodthewas
experimental results,
the second-best wehad
and canasee that ability
certain the error of the
to fight EKF algorithm
against under
noise. However,
different noises was within 2%, which would meet the practical requirements.
the incremental seeking method exhibited the poorest performance, with the highest error, The dichot-
omous to
failing method was the second-best
meet practical requirementsand andhad a certain
requiring abilityimprovements.
further to fight against noise. How-
ever,With
the incremental seeking method exhibited the poorest performance,
regard to the impact of voltage noise on the algorithm’s estimation with the highest
performance,
error, failing to meet practical requirements and requiring further improvements.
the EKF algorithm effectively filtered out the noise and was almost unaffected. In contrast,
both With regard to the
the incremental impactand
seeking of voltage noise on
dichotomous the algorithm’s
methods were more estimation performance,
susceptible to voltage
the EKF algorithm effectively filtered out
noise, resulting in lower stability and [Link] noise and was almost unaffected. In contrast,
Batteries 2023, 9, 442 17 of 20

4.3. Algorithm Computational Efficiency

Because the process of completing the overall battery SOC estimation was long, the
computational efficiency of the three algorithms was considered by comparing the time
required to perform a single SOC estimation among them.
The algorithms were run in the following environment: under PC conditions using
processor: Intel (R) Core (TM) i5-10200H CPU @ 2.40 GHz, RAM configuration:16.00 GB,
and MATLAB software.
The time consumption of all algorithms is shown in Table 6. The incremental pathfind-
ing method took 0.024 s and 0.025 s to complete a single SOC estimation without a current
sensor under the two operating conditions. The dichotomous method took 0.018 s and
0.02 s. The EKF algorithm took 0.012 s and 0.007 s. The ratio of the calculated efficiencies of
the three methods at DST conditions is about 2:3:4.

Table 6. Algorithm operation schedule.

Algorithm
Incremental Seeking Method Dichotomous Method EKF
Work Conditions
Total time 331 s 250 s 171 s
DST Average data estimation time
0.024 s 0.018 s 0.012 s
per frame
Total time 636 s 498 s 191.0 s
HPPC Average data estimation time
0.025 s 0.020 s 0.007 s
per frame

4.4. Repeatability Verification

In order to avoid the chance of single battery data, more test data of LCO lithium-ion
batteries were used to verify the reliability of the algorithm. The HPPC operation data of
more batches of batteries were used to verify the algorithm. The average error is shown in
Table 7.

Table 7. Estimation error at different initial SOC states in HPPC conditions.

Accurate Initial SOC Initial SOC 0.9 Initial SOC 0.7

Incremental
soc Voltage soc Voltage soc Voltage
seeking method
Average value 7.59% 0.026 V 7.74% 0.030 V 9.16% 0.041 V
Maximum value 15.72% 0.336 V 15.21% 0.339 V 30% 0.697 V
Dichotomous
soc Voltage soc Voltage soc Voltage
method
Average value 3.23% 0.026 V 3.41% 0.029 V 5.42% 0.037 V
Maximum value 7.73% 0.336 V 10.00% 0.287 V 30% 0.276 V
EKF soc Voltage soc Voltage soc Voltage
Average value 3.06% 0.010 V 3.67% 0.010 V 4.73% 0.011 V
Maximum value 9.07% 0.308 V 11.22% 0.4003 V 30.16% 0.4003 V

In order to ensure the reliability of the algorithm, the simulation results of constant
current charge-discharge test data of the LFP lithium-ion battery are also presented. When
the initial SOC is accurate, the average error results of the EKF algorithm are shown in
Table 8.
Batteries 2023, 9, 442 18 of 20

Table 8. Average error of EKF algorithm for eight batteries.

Battery Number A B C D E F G H
Batteries 2023, 9, x FOR PEER REVIEW 18 of 20
SOC 1.99% 1.67% 1.08% 1.76% 1.73% 1.52% 1.46% 1.56%
Voltage 0.0185 V 0.0168 V 0.0203 V 0.0189 V 0.0177 V 0.173 V 0.0180 V 0.0180 V

The SOC estimation results of the battery numbered A are shown in Figure 17.
The SOC estimation results of the battery numbered A are shown in Figure 17.

Figure 17. Estimation results of the EKF algorithm for HPPC conditions.

When using
When using different
different battery
battery test
test data,
data, the
the EKF
EKF algorithm
algorithm can
can still
still maintain
maintain aa good
good
estimation effect.
estimation effect.
5. Conclusions
5. Conclusions
It is important to obtain current information for accurate SOC estimation of lithium-ion
It is important to obtain current information for accurate SOC estimation of lithium-
batteries. However, due to concerns about cost, size, and power consumption, current
ion batteries. However, due to concerns about cost, size, and power consumption, current
sensors are sometimes not equipped in portable devices.
sensors are sometimes not equipped in portable devices.
In this work, three methods were developed for lithium battery SOC estimation that
In this work, three methods were developed for lithium battery SOC estimation that
provide new solutions for SOC estimation for small portable devices in the absence of
provide new solutions
current monitoring and for SOCgap
fill the estimation
in existing forstudies.
small portable devices inand
The dichotomous theincremental
absence of
current
seeking monitoring
methods were andfound
fill thetogap in existing
be sensitive to studies. Theaccuracy,
initial SOC dichotomous and incremental
and estimation errors
seeking methods were found to be sensitive to initial SOC accuracy,
exceeded the allowed range. For instance, using the dichotomous method led to a 6.37% and estimation errors
er-
exceeded the allowed
ror at an initial SOC of range.
70% underFor instance, using the The
HPPC conditions. dichotomous
incremental method
seeking ledmethod
to a 6.37%
was
error at ansusceptible
also more initial SOC to of voltage
70% under noise HPPC
than conditions.
the other two Themethods,
incrementalwithseeking
a 10 mV method
noise
was also more susceptible to voltage noise
causing an estimation error of 5.92% in DST conditions. than the other two methods, with a 10 mV noise
causing
Thean estimation error
dichotomous methodof 5.92%
and in theDST conditions.
extended Kalman filter method demonstrated
superior accuracy in the presence of certain measurementfilter
The dichotomous method and the extended Kalman method
noises, with demonstrated
estimation errors su-
perior accuracy in the presence of certain measurement noises,
under 100 mV noise controlled within ±2%. The extended Kalman filter algorithm was with estimation errors un-
der 100 mV effective
particularly noise controlled
in filteringwithin
out the ±2%. Thewith
noise, extended Kalman filter
error variation algorithm
following noisewas par-
addition
ticularly effective in filtering out the noise, with error variation
not exceeding 0.04%. Additionally, the extended Kalman filter method outperformed the following noise addition
not
otherexceeding
two methods 0.04%. inAdditionally,
terms of estimationthe extended
speed, Kalman filter method
with an average time ofoutperformed
0.01 s per framethe
other
of datatwo methods in terms of estimation speed, with an average time of 0.01 s per frame
estimated.
of data [Link] is worth noting that the estimation error of SOC will be larger when the
In addition,
In addition,
initial SOC error it
is is worth
large duenoting
to the that the estimation
absence error of
of current input. SOC willmeasures
However, be largercanwhenbe
the
taken to avoid excessive errors. For example, the battery is left for a period of time, and can
initial SOC error is large due to the absence of current input. However, measures the
be takenistocalibrated
battery avoid excessive
by OCV errors.
after For
it is example,
fully [Link] battery
In thatiscase,
left for a period
this weakness of time,
doesand
not
the battery
affect is calibrated
the usefulness andby OCV after itof
sophistication is the
fully charged. In
algorithms that case, this weakness does
used.
not affect the usefulness
Considering the aging and sophistication
process of thethe
of the battery, algorithms [Link] will change,
battery model
leading to an increase
Considering in SOCprocess
the aging estimationof theerror. In order
battery, the to solve this
battery problem,
model a possible
parameters will
change, leading to an increase in SOC estimation error. In order to solve this problem, a
possible approach is to use an improved extended Kalman filter (IEKF) algorithm [32] or
a model adaptive EKF (MAEKF) [33] algorithm. The general idea is to select the parame-
ters of the electrical model with high sensitivity. In view of the fact that the battery voltage
Batteries 2023, 9, 442 19 of 20

approach is to use an improved extended Kalman filter (IEKF) algorithm [32] or a model
adaptive EKF (MAEKF) [33] algorithm. The general idea is to select the parameters of the
electrical model with high sensitivity. In view of the fact that the battery voltage derivative
changes abruptly twice with time when the discharge current is constant, the parameters
are updated using IEKF or MAEKF. Thus, the excellent accuracy of the model is maintained
during the aging process of the battery.

Author Contributions: Conceptualization, J.L.; methodology, J.L.; software, J.L.; validation, Z.Z.;
formal analysis, Z.Z.; investigation, Z.Z.; resources, J.L.; data curation, Z.Z.; writing—original draft
preparation, Z.Z.; writing—review and editing, J.S., J.L., Y.W.; visualization, Z.Z.; supervision,
Z.W.; project administration, J.L. All authors have read and agreed to the published version of
the manuscript.
Funding: This research received no external funding.
Data Availability Statement: No new data were created or analyzed in this study. Data sharing is
not applicable to this article.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Kulova, T.L.; Fateev, V.N.; Seregina, E.A.; Grigoriev, A.S. A Brief Review of Post-Lithium-Ion Batteries. Int. J. Electrochem. Sci.
2020, 15, 7242–7259. [CrossRef]
2. Garcia Elvira, D.; Machado, R.; Plett, G.L.; Trimboli, M.S.; Valderrama Blavi, H.; Cid Pastor, A.; Martinez Salamero, L. Simplified
Li Ion Cell Model for BMS Coupling an Equivalent Circuit Dynamic Model with a Zero Dimensional Physics Based SEI Model.
J. Electrochem. Soc. 2021, 168, 110526. [CrossRef]
3. Miron-Alexe, V. Portable lithium-ion battery ups with bms function for raspberry pi and other iot embedded systems. J. Sci. Arts
2022, 22, 763–780. [CrossRef]
4. Wang, C.; Liu, Z.; Sun, Y.; Gao, Y.; Yan, P. Aging Behavior of Lithium Titanate Battery under High-Rate Discharging Cycle.
Energies 2021, 14, 5482. [CrossRef]
5. Wu, L.; Liu, K.; Pang, H.; Jin, J. Online SOC Estimation Based on Simplified Electrochemical Model for Lithium-Ion Batteries
Considering Current Bias. Energies 2021, 14, 5482. [CrossRef]
6. Li, J.; Zhao, M.; Dai, C.; Wang, Z.; Pecht, M. A mathematical method for open-circuit potential curve acquisition for lithium-ion
batteries. J. Electroanal. Chem. 2021, 895, 115488. [CrossRef]
7. Orazem, M.E. Electrochemical impedance spectroscopy: The journey to physical understanding. J. Solid State Electrochem. 2020,
24, 2151–2153. [CrossRef]
8. Xing, L.; Zhan, M.; Guo, M.; Ling, L. Parameter identification and SOC estimation for power battery based on multi-timescale
double Kalman filter algorithm. Int. J. Comput. Sci. Eng. 2022, 25, 619–628. [CrossRef]
9. Fang, C.; Jin, Z.; Wu, J.; Liu, C. Estimation of Lithium-Ion Battery SOC Model Based on AGA-FOUKF Algorithm. Front. Energy
Res. 2021, 9, 769818. [CrossRef]
10. Huang, Z.; Fang, Y.; Xu, J. Soc estimation of li-ion battery based on improved ekf algorithm. Int. J. Automot. Technol. 2021, 22,
335–340. [CrossRef]
11. Xing, L.; Wu, X.; Ling, L.; Lu, L.; Qi, L. Lithium Battery SOC Estimation Based on Multi-Innovation Unscented and Fractional
Order Square Root Cubature Kalman Filter. Appl. Sci. 2022, 12, 9524. [CrossRef]
12. Chen, L.; Wu, X.; Tenreiro Machado, J.A.; Lopes, A.M.; Li, P.; Dong, X. State-of-Charge Estimation of Lithium-Ion Batteries Based
on Fractional-Order Square-Root Unscented Kalman Filter. Fractal Fract. 2022, 6, 52. [CrossRef]
13. Wang, Q.; Ye, M.; Wei, M.; Lian, G.; Li, Y. Deep convolutional neural network based closed-loop SOC estimation for lithium-ion
batteries in hierarchical scenarios. Energy 2023, 263, 125718. [CrossRef]
14. Cui, Z.; Kang, L.; Li, L.; Wang, L.; Wang, K. A hybrid neural network model with improved input for state of charge estimation of
lithium-ion battery at low temperatures. Renew. Energy 2022, 198, 1328–1340. [CrossRef]
15. Chen, N.; Zhao, X.; Chen, J.; Xu, X.; Zhang, P.; Gui, W. Design of a Non-Linear Observer for SOC of Lithium-Ion Battery Based on
Neural Network. Energies 2022, 15, 3835. [CrossRef]
16. Zhou, Y.; Wang, S.; Xie, Y.; Zhu, T.; Fernandez, C. An improved particle swarm optimization-least squares support vector
machine-unscented Kalman filtering algorithm on SOC estimation of lithium-ion battery. Int. J. Green Energy 2023. [CrossRef]
17. Manoharan, A.; Begam, K.M.; Aparow, V.R.; Sooriamoorthy, D. Artificial Neural Networks, Gradient Boosting and Support Vector
Machines for electric vehicle battery state estimation: A review. J. Energy Storage 2022, 55, 105384. [CrossRef]
18. Wu, X.; Mi, L.; Tan, W.; Qin, J.; Zhao, M. State of Charge (SOC) Estimation of Ni-MH Battery Based on Least Square Support
Vector Machines. In Proceedings of the International Conference on Mechatronics and Intelligent Materials, Lijiang, China, 21–22
May 2011; pp. 1204–1209.
Batteries 2023, 9, 442 20 of 20

19. Xuan, L.; Qian, L.; Chen, J.; Bai, X.; Wu, B. State-of-Charge Prediction of Battery Management System Based on Principal
Component Analysis and Improved Support Vector Machine for Regression. IEEE Access 2020, 8, 164693–164704. [CrossRef]
20. Xu, X.; Yue, X.; Chen, Y.; Liang, Z. Li Plating Regulation on Fast-Charging Graphite Anodes by a Triglyme-LiNO3 Synergistic
Electrolyte Additive. Angew. Chem. Int. Ed. 2023, 62, e202306963. [CrossRef] [PubMed]
21. He, J.; Meng, S.; Yan, F. A Comparative Study of SOC Estimation Based on Equivalent Circuit Models. Front. Energy Res. 2022,
10, 914291. [CrossRef]
22. Ruan, H.; Sun, B.; Zhang, W.; Su, X.; He, X. Quantitative Analysis of Performance Decrease and Fast-Charging Limitation for
Lithium-Ion Batteries at Low Temperature Based on the Electrochemical Model. IEEE Trans. Intell. Transp. Syst. 2021, 22, 640–650.
[CrossRef]
23. Wu, S.; Pan, W.; Zhu, M. A Collaborative Estimation Scheme for Lithium-Ion Battery State of Charge and State of Health Based
on Electrochemical Model. J. Electrochem. Soc. 2022, 169, 090516. [CrossRef]
24. Wang, J.; Zhang, L.; Xu, D.; Zhang, P.; Zhang, G. A Simplified Fractional Order Equivalent Circuit Model and Adaptive Online
Parameter Identification Method for Lithium-Ion Batteries. Math. Probl. Eng. 2019, 2019, 6019236. [CrossRef]
25. Cambron, D.C.; Cramer, A.M. A Lithium-Ion Battery Current Estimation Technique Using an Unknown Input Observer. IEEE
Trans. Veh. Technol. 2017, 66, 6707–6714. [CrossRef]
26. Putra, W.S.; Dewangga, B.R.; Cahyadi, A.; Wahyunggoro, O.; IEEE. Current Estimation Using Thevenin Battery Model. In
Proceedings of the Joint International Conference On Electric Vehicular Technology and Industrial, Mechanical, Electrical, and
Chemical Engineering (ICEVT & IMECE) 2015, Surakarta, Indonesia, 4–5 November 2015; pp. 5–9.
27. Chun, C.Y.; Baek, J.; Seo, G.S.; Cho, B.H.; Kim, J.; Chang, I.K.; Lee, S. Current sensor-less state-of-charge estimation algorithm for
lithium-ion batteries utilizing filtered terminal voltage. J. Power Sources 2015, 273, 255–263. [CrossRef]
28. Hou, J.; Yang, Y.; Gao, T. A Variational Bayes Based State-of-Charge Estimation for Lithium-Ion Batteries without Sensing Current.
IEEE Access 2021, 9, 84651–84665. [CrossRef]
29. Pei, M.; Shi, H.; Yao, F.; Liang, S.; Xu, Z.; Pei, X.; Wang, S.; Hu, Y. 3D printing of advanced lithium batteries: A designing strategy
of electrode/electrolyte architectures. J. Mater. Chem. A 2021, 9, 25237–25257. [CrossRef]
30. Li, J.; Wang, L.; Lyu, C.; Liu, E.; Xing, Y.; Pecht, M. A parameter estimation method for a simplified electrochemical model for
Li-ion batteries. Electrochim. Acta 2018, 275, 50–58. [CrossRef]
31. Liu, Y.; Tang, S.; Li, L.; Liu, F.; Jiang, L.; Jia, M.; Ai, Y.; Yao, C.; Gu, H. Simulation and parameter identification based on
electrochemical-thermal coupling model of power lithium ion-battery. J. Alloys Compd. 2020, 844, 156003. [CrossRef]
32. Sepasi, S.; Ghorbani, R.; Liaw, B.Y. Improved extended Kalman filter for state of charge estimation of battery pack. J. Power Sources
2014, 255, 368–376. [CrossRef]
33. Sepasi, S.; Ghorbani, R.; Liaw, B.Y. A novel on-board state-of-charge estimation method for aged Li-ion batteries based on model
adaptive extended Kalman filter. J. Power Sources 2014, 245, 337–344. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.