Article

A Comparative Study of Electric Vehicles Battery State of Charge

Estimation Based on Machine Learning and Real Driving Data
Salma Ariche 1,2, * , Zakaria Boulghasoul 2 , Abdelhafid El Ouardi 1, * , Abdelhadi Elbacha 2 , Abdelouahed Tajer 2
and Stéphane Espié 3

1 Université Paris-Saclay, ENS Paris-Saclay, CNRS, SATIE, 91110 Gif-sur-Yvette, France

2 Engineering Systems and Applications (LISA), Cadi Ayyad University, Marrakech 40000, Morocco;
[email protected] (Z.B.); [email protected] (A.E.); [email protected] (A.T.)
3 Université Gustave Eiffel, SATIE, 91190 Gif-sur-Yvette, France; [email protected]
* Correspondence: [email protected] (S.A.);
[email protected] (A.E.O.)

Abstract: Electric vehicles (EVs) are rising in the automotive industry, replacing combustion engines
and increasing their global market presence. These vehicles offer zero emissions during operation and
more straightforward maintenance. However, for such systems that rely heavily on battery capacity,
precisely determining the battery’s state of charge (SOC) presents a significant challenge due to its
essential role in lithium-ion batteries. This research introduces a dual-phase testing approach, consid-
ering factors like HVAC use and road topography, and evaluating machine learning models such as
linear regression, support vector regression, random forest regression, and neural networks using
datasets from real-world driving conditions in European (Germany) and African (Morocco) contexts.
The results validate that the proposed neural networks model does not overfit when evaluated using
the dual-phase test method compared to previous studies. The neural networks consistently show
high predictive precision across different scenarios within the datasets, outperforming other models
by achieving the lowest mean squared error (MSE) and the highest R² values.

Keywords: battery electric vehicles; state of charge; online estimation; real driving data; machine
Citation: Ariche, S.; Boulghasoul, Z.;
learning; neural networks
El Ouardi, A.; Elbacha, A.; Tajer, A.;
Espié, S. A Comparative Study of
Electric Vehicles Battery State of
Charge Estimation Based on Machine
Learning and Real Driving Data. J.
1. Introduction
Low Power Electron. Appl. 2024, 14, 59. Today’s automotive sector is moving towards intelligent transportation, which is
https://doi.org/10.3390/jlpea14040059 emerging as a pivotal global industry, wielding significant influence not only on the
economic front but also in the realm of research and development. A growing array of
Academic Editor: Giorgos
Dimitrakopoulos
technological components are being integrated into vehicles to enhance safety and comfort.
In light of the critical imperative to combat climate change and lower air pollution
Received: 29 October 2024 levels [1], the environmental benefits of using electric vehicles (EV) are especially sig-
Revised: 26 November 2024 nificant. As governments and policymakers worldwide intensify their efforts to meet
Accepted: 9 December 2024 these challenges, promoting EVs has become a transition towards decarbonization of the
Published: 11 December 2024
transportation system since manufacturers are electrifying their fleets to meet new legal
requirements. The shift to EVs is seen as a technical advancement and a crucial step to-
ward sustainable transportation. Advances in battery technology, charging infrastructure,
Copyright: © 2024 by the authors.
and regulations supporting low-emission cars contribute to this change [2]. However,
Licensee MDPI, Basel, Switzerland. despite the increased supply of EVs and improving infrastructure related to EV charging,
This article is an open access article vehicle registration statistics indicate that customers still need to embrace EVs [3].
distributed under the terms and The EV is characterized by its low noise emissions, user-friendly operation, and ab-
conditions of the Creative Commons sence of fuel-related expenses commonly associated with traditional vehicles. Despite
Attribution (CC BY) license (https:// these advancements, their limited operating range remains a significant barrier to the
creativecommons.org/licenses/by/ broad acceptance of battery electric vehicles (BEVs). The reluctance to fully embrace BEVs
4.0/). is partly due to current battery technology, which is rapidly evolving yet still does not

J. Low Power Electron. Appl. 2024, 14, 59. https://doi.org/10.3390/jlpea14040059 https://www.mdpi.com/journal/jlpea

J. Low Power Electron. Appl. 2024, 14, 59 2 of 16

match the energy density and recharging speed of traditional gasoline engines [4]. Thus,
the precise assessment of the SOC of batteries has become indispensable for drivers and
the energy regulation systems of these vehicles. The SOC is a vital indicator of a battery’s
residual charge, which not only informs the potential driving range before the necessity
of recharging arises but also aids in mitigating risks associated with overcharging and
undercharging [5]. It is similar to the gasoline gauge in traditional automobiles but is more
complex to estimate accurately due to the nature of battery chemistries [6]. Over the last
decades, extensive scientific research has focused on determining the SOC of batteries. Ow-
ing to the complex and non-linear electrochemical properties of batteries, their performance
hinges on a variety of factors, both intrinsic and environmental. Methods for estimating
the SOC leverage available data through various techniques, including open circuit voltage
(OCV), the Kalman filter (KF), and artificial intelligence (AI) approaches. It is also critical to
evaluate not just individual battery performance but the collective performance of the entire
battery assembly. Moreover, challenges like battery degradation, lifespan, temperature
variations, among others, present considerable difficulties in precisely gauging the SOC [7].
In the past few years, machine learning (ML) has become an essential tool in many ar-
eas, including the automotive industry [8,9] and energy management systems (EMS) [10,11].
Using machine learning to estimate the SOC of BEVs is an important step forward in solv-
ing the problem of range anxiety. ML algorithms can learn from driving data, patterns of
battery use, and environmental factors. They provide a more accurate and reliable SOC
estimation than traditional methods. This enhancement is critical in improving the user
experience by providing more precise range predictions and a more comfortable ride, thus
fostering greater confidence and adoption of EVs among potential users. ML used in SOC
estimations opens up smart energy management possibilities for BEVs. By accurately
determining the SOC, the EMS can use the battery best, improve the vehicle’s performance,
and even make the battery last longer through efficient charging and discharging strate-
gies [12]. Therefore, precise SOC estimation not only contributes to the user’s convenience
but also the vehicle’s overall safety. By incorporating ML-based SOC estimation into in-
telligent transportation systems (ITS), BEVs can establish communication with charging
stations, traffic management systems, and other vehicles in order to optimize traffic flow,
charging schedules, and driving routes. The research work presented in the current paper
focuses on the estimation of the battery electric vehicle SOC using data-driven approaches
like regression models and artificial neural networks, as well as trained real-world driving
data derived from two different BEVs.
The rest of this paper is organized as follows: Section 2 provides a comprehensive
review of current methods for SOC estimation. Section 3 details the methodology we
propose. Section 4 discusses the experimental results. The paper concludes with Section 5,
which offers conclusions and future outlooks of this study.

2. Related Work
In general, battery SOC is characterized as the fraction of the remaining energy relative
to the total energy capacity it can store (1). With Qc representing the residual energy and Q0
denoting the initial energy available under specific conditions, Q T is the amount of energy
that the battery has previously discharged. The SOC can, therefore, be mathematically
expressed using these variables.

Qc Q
SOC = = 1− T (1)
Q0 Q0

However, this mathematical representation of the battery’s available capacity does

not fully capture its practical variability, as SOC is frequently affected by a variety of
conditions [13]. The battery SOC can be calculated using several methods, the most
common of which involve the integration of current over time and voltage-based models.
The methods for estimating SOC can be categorized into three groups:
J. Low Power Electron. Appl. 2024, 14, 59 3 of 16

a. Conventional Methods
Approaches that heavily depend on battery parameters, such as the straightforward
look-up table methods [14], are primarily represented by the open-circuit voltage
(OCV), which assumes that there is a linear relation between the voltage, the SOC and
the ampere-hour (Ah) integration method. Yet, these methods exhibit a substantial
estimation error and may not be well-suited for real-time estimation across different
battery types. The association between the SOC and the open-circuit voltage (OCV)
depends on the battery’s type and capacity. A lead-acid battery, for instance, shows a
direct linear relationship between SOC and OCV, unlike lithium-ion batteries with
more complex and non-linear connections [15].
b. Model Based Methods
Model-based approaches were presented to overcome the limitations of traditional
methods that rely solely on battery parameters. The main goal of model-based SOC
estimation lies in utilizing the battery model to establish a connection between battery
signals. Model-based techniques such as the Electrochemical Impedance Model (EIM)
and electrochemical impedance spectroscopy (EIS) [16,17] are used to discern the
relationship between a battery’s impedance and its state of charge (SOC). Alternatively,
linking the model to cellular measurements through the use of a filtering algorithm
derived from current control theory can establish a closed-loop system for estimation.
This method capitalizes on the variances between the modeled and actual battery
outputs to refine the state predictions, thereby reducing the discrepancy between
the modeled battery and the measured signals. The Kalman filter (KF) is a popular
filter-based model; however, it is applicable only to linear systems. The extended
Kalman filters (EKF) and Sigma-Point Kalman filters (SPKF) extend this capability
to nonlinear systems [18–20]. The KF and its variants operate on the premise that all
noises are characterized by zero mean and follow a Gaussian distribution. In contrast,
the adaptive Kalman filter (AKF) estimates both the model and the noises online, while
the adaptive extended Kalman filter (AEKF) is utilized to enhance state estimation
accuracy [21].
c. Data Driven Methods
The inaccuracy of SOC estimation using model-based or traditional methods is still
high when comparing the estimated output and the actual SOC recorded in the
datasets. Furthermore, they are methods that are time-consuming and have low
efficiency. Data-driven methods are established to accurately estimate the SOC based
on recorded data; we are talking about artificial intelligence (AI) algorithms that take
information from large datasets for model training [22]. Machine learning (ML) mod-
els forecast battery SOC based on time-series datasets, including features that could
impact the SOC over time. SOC estimation often employs regression algorithms like
linear regression (LR), support vector machine (SVM), support vector regressor (SVR),
Gaussian process regression (GPR), random forest regression (RF), and decision tree
regression (DT), as documented in various references [23–26]. The works of [27,28] in-
troduced a hybrid SOC estimation approach combining model-based and data-driven
methods for better accuracy. Deep learning (DL) or neural networks (NNs) in research
have been shown to give better results related to SOC estimation [29]. An artificial
neural network (ANN) model is created to estimate the SOC accurately. The esti-
mation relies on load classification, which includes post-processing and boosting
techniques to prevent overfitting, as discussed in [30]. Additionally, a comparative
study of DL algorithms for lithium-ion SOC estimation was carried out in the works
of [31,32]. The papers [33–35] also introduce the use of a long short-term memory
recurrent neural network (LSTM-RNN) for SOC estimation. In [36], a deep multilayer
perceptron network was developed and assessed using different activation functions,
including the Sigmoid, tanh, ReLU, and Mish activation functions. Additional stud-
ies have focused on analyzing the effects of HVAC on electric vehicles, distributed
power generation, and electrical storage using AI for accurate SOC estimation and to
J. Low Power Electron. Appl. 2024, 14, 59 4 of 16

improve range and battery lifetime [37]. Statistical analysis using ARIMA models and
LTSM was dedicated to SOC forecasting in micro-grid systems [38].
While offering enhanced accuracy and adaptability in SOC estimation, data-driven
methods also present several challenges. One significant drawback is their depen-
dency on vast and high-quality datasets, which may not always be available or may
require extensive preprocessing to be valid. These methods can also be computa-
tionally intensive, leading to longer processing times and requiring more powerful
hardware, which can increase the cost and complexity of implementation. These
methods may need to be more balanced, especially in scenarios where the variability
in the training data do not accurately reflect real-world conditions, which can lead the
models to overfit.
In this research, we evaluate the batteries’ state of charge (SOC) estimation using
real-world driving data, a notable departure from the controlled laboratory conditions in
the existing literature. To comprehensively understand EV energy consumption, we under-
score the importance of incorporating diverse driving scenarios and auxiliary components,
such as HVAC systems, weather conditions, and specific route information. By leveraging
datasets from European (Germany) and African (Morocco) contexts, we address the com-
plexities of real-world EV operation, particularly highlighting the underexplored challenges
of African road conditions. Furthermore, we compare four machine learning models and
introduce a second testing phase to combat overfitting, ensuring model robustness and
reliability across various driving conditions. This dual-phase testing enhances the practical
applicability of our models, marking a significant contribution to the field.

3. Methodology
In a previous study [39], the paper analyzed how the EV’s energy consumption could
be affected by factors like road topography and ambient temperature when switching to
comfort ride mode on Moroccan roads. However, the evaluation was insufficient for an
accurate BEV range estimation in a real-life scenario, whether for a regular or a comfort ride
option. Building upon the EV model in that study, the research was extended by evaluating
the SOC estimation, including two crucial energy consumers as presented in Figure 1.

Number Outside Target

of Passengers Temperature Temperature

Vehicle’s velocity
Thermal Management
System (TMS)
Acceleration Air Conditioning,
commands Power Converter Motor Ventilation
H-Bridge with Permanent
Driver and Heating
regenerative magnet DC
Controller
breaking Node 2: TMS
motor
Breaking
commands
Node 1: Powertrain
Vehicle Body
Energy to power nodes
Drive cycle
SOC
Reference velocity Energy recovered by regenerative braking Battery State of
Lithium/Ion Charge Estimation
Battery

Figure 1. BEV model with powertrain and thermal management system.

The block diagram illustrates an electric vehicle’s system, where the driver controller
interprets the driver’s input via the drive cycle to manage the vehicle’s speed. The power-
train includes a power converter responsible for modulating the power from the battery to
drive the motor, which in turn propels the vehicle. Additionally, the thermal management
system (TMS) from [40–42], regulates the temperature for optimal performance and comfort,
encompassing the HVAC system that handles air conditioning, ventilation, and heating
within the vehicle’s cabin and its components. The TMS takes as input parameters the target
J. Low Power Electron. Appl. 2024, 14, 59 5 of 16

cabin temperature, the outside temperature, the number of passengers, and the vehicle’s
speed. Finally, the lithium-ion type battery is central to the system, providing energy to
the powertrain and TMS and recovering energy through regenerative braking. The battery
block uses KF and EKF for SOC estimation. The estimation results using the model-based
method were not as satisfying as expected, with a significant difference when comparing
the simulation output and the actual recorded SOC in the datasets. This difference might be
explained by several limits related to the MATLAB/SIMULINK existing model adaptation,
as listed below:
(i) Model Accuracy: It can sometimes encourage idealized modeling that does not take
into account all the non-linearities and uncertainties present in real systems.
(ii) Documentation: It may sometimes be insufficient, particularly for advanced function-
alities or custom blocks.
(iii) Limited Customization: In some cases, it is impossible to introduce input parameters.
(iv) Simulation parameters: Managing simulation parameters to guarantee test repro-
ducibility is a challenge when using real datasets.
(v) Performance: The simulated BEV is a model with a large number of Simulink and
Simscape blocks, making simulation time-consuming and consuming a lot of sys-
tem resources.
As an alternative approach, in this section, we present the chosen data-driven methods
used to estimate the battery SOC, tested on actual data from a Moroccan and European
context. The models were trained and tested with Python 3.10.12 version using ML li-
braries like scikit-learn (sklearn), Tensorflow, and Keras [43–45]. The flow chart in Figure 2
illustrates the process of SOC forecasting using ML algorithms.

Figure 2. Flow chart of the proposed data-driven methods to estimate battery SOC.
J. Low Power Electron. Appl. 2024, 14, 59 6 of 16

Initially, data preprocessing is imperative to refine the collected data, involving clean-
ing to remove noise or irrelevant information, transformation to ensure consistency, and fea-
ture selection to identify the most relevant parameters for SOC estimation. Selecting appro-
priate features is crucial as it directly influences the model’s accuracy and computational
efficiency. The choice of algorithms—linear regression (LR), support vector regression
(SVR), random forest (RF), and neural networks (NNs)—strikes a balance between sim-
plicity and precision. LR, with its straightforward computation, offers a baseline for
performance. SVR was selected because of its capacity to handle non-linear relationships.
RF is included for its robustness to overfitting and its ability to rank the importance of
features. At the same time, NNs were chosen because of their high potential for accuracy in
complex relationships. Performance indices like mean absolute error (MAE), mean squared
error (MSE), root mean squared error (RMSE), and R² are utilized to evaluate the models
rigorously on an initial and second test sets. These metrics give us a comprehensive view
of the model’s accuracy and the variance explained by the model [46]. The second test
set is additional recorded trips with the same number of features. Including a secondary
test set, sourced from different operational contexts, is a strategic measure to assess the
generalizability of the models and safeguard against overfitting. This dual-phase testing
ensures that the algorithms perform well across various scenarios, which is essential for
reliable SOC estimation in the dynamic environment of electric vehicles. The algorithms
enlisted for SOC estimation in battery electric vehicles are detailed as follows:
1. Linear Regression: A statistical model that describes a linear relationship between
dependent and independent variables. LR is a supervised ML algorithm applied to
models with continuous linear variable output. It is often used for its simplicity and
performance when tested on small- or medium-sized datasets. The empirical equation
of the model is:
y = β 0 + β 1 x1 + β 2 x2 + · · · + β n x n + ϵ (2)
where y is the predicted variable, x1 , x2 , . . . , xn are the input feature values, β0 , β1 , . . . , βn
are the coefficients, and ϵ is the error term as presented in Algorithm 1.

Algorithm 1: Linear Regression

Input: Dataset with predictor variable x and target variable y
Output: Model coefficients β 0 and β 1
Define the model: y = β 0 + β 1 x + ϵ;
Estimate coefficients using least squares;
Evaluate the model using R-squared, MSE, and MAE;
Make predictions: Use the model to predict y for new x values;

2. Support Vector Regression: SVR, or support vector regressor, is a supervised learning

algorithm employed for addressing various nonlinear classification and regression
challenges. It operates on a principle similar to the support vector machine (SVM) but
focuses on predicting the value of the output rather than classifying it into categories.
As in Algorithm 2, SVR tackles regression problems by identifying a hyperplane within
a high-dimensional feature space that optimally captures the connection between
input variables (features) and the target variable (output). It proves to be a robust
tool for regression tasks, particularly when dealing with datasets characterized by
intricate relationships between variables.
J. Low Power Electron. Appl. 2024, 14, 59 7 of 16

Algorithm 2: Support Vector Regression (SVR)

Input: Training data, chosen kernel function, hyperparameters C and ϵ
Output: SVR model for prediction
Select kernel function (e.g., linear, polynomial, RBF);
Define the SVR model with the chosen kernel and hyperparameters C and ϵ;
Formulate the optimization problem to find the hyperplane that minimizes error
within the ϵ margin;
Solve the optimization problem;
Identify support vectors;
Compute predictions;
Apply inverse transformations to predictions if needed;
Evaluate model performance using regression metrics;
Fine-tune hyperparameters (e.g., C and ϵ) using cross-validation;

3. Random Forest regression: An ensemble learning strategy. Algorithm 3 functions by

generating several decision trees during the training process and deriving a collective
prediction from the outputs of these individual trees. This technique focuses on the
aggregated knowledge obtained from various trees instead of depending on a solitary
model, thereby achieving a more robust and more dependable prediction through the
mean of the outcomes from every tree that forms part of the ensemble.

Algorithm 3: Random Forest Regression

Input: Training data
Output: Ensemble model for prediction
Bootstrap Sampling: Create multiple random subsets of the data;
Build Decision Trees: Grow decision trees using different subsets of features;
Aggregate Predictions: Average the predictions from all decision trees;
Predict new data points using the ensemble model;

4. Neural Networks (NNs): These models are used to describe a non-linear or complex
interaction using a set of input and output data. They consist of layers of intercon-
nected nodes where each node performs a simple computation. The connections
between neurons have associated weights tuned during the learning process as de-
scribed in Algorithm 4. Different NN architectures exist in the literature, such as:
(a) Feedforward Neural Networks (FNNs): The simplest architecture of NNs is
a straightforward connection between the NN nodes. It includes the basic
perceptron and multi-layer perceptrons (MLPs).
(b) Convolutional Neural Networks (CNNs): Based on convolution operations for
data processing as a matrix-like topology, they are primarily used in image
processing and computer vision.
(c) Recurrent Neural Networks (RNNs): Unlike FNNs, RNNs have connections
that form cycles between nodes. These cyclical connections are designed to re-
tain a form of memory, making RNNs particularly well-suited for applications
involving sequential data. It includes, but is not limited to, time series analysis
and natural language processing tasks.
(d) Long Short-Term Memory (LSTM) Networks: A specialized variant of RNNs
designed to learn and retain long-term dependencies in data sequences.
J. Low Power Electron. Appl. 2024, 14, 59 8 of 16

Algorithm 4: Artificial Neural Network (ANN)

Input: Training data with input and output variables
Output: Trained neural network model for predictions
Design Network Structure: Define layers and neurons;
Initialize weights and biases;
foreach epoch do
Forward Propagation: Pass data through the network;
Calculate Loss: Compute the difference between output and actual values;
Backpropagation: Adjust weights and biases based on the loss;
Evaluate model performance on validation data;
Make predictions on new data;

The chosen PIs for the algorithm evaluation are listed in Table 1.

Table 1. Overview of Error Metrics.

Error Metric Description Equation

1
MAE (Mean Absolute A metric that measures the difference between two variables that MAE = n ∑ |yi − ŷi |
Error) are continuous. It is determined by computing the mean of the
absolute variations between the forecasted values and the actual
observed data.
Where yi is the actual value, ŷi is the predicted value, and n is the
number of observations.
1
MSE (Mean Squared Computes the average of the squared differences, representing the MSE = n ∑(yi − ŷi )2
Error) variance between the predicted and actual values. Due to the squar-
ing of errors, MSE is more sensitive to outliers compared to the MAE.
Where yi is the actual value, ŷi is the predicted value, and n is the
number of observations.
q
1
RMSE (Root Mean Quantifies the average magnitude of residuals, emphasizing larger RMSE = n ∑(yi − ŷi )2
Squared Error) errors more significantly. RMSE is particularly useful in scenarios
where large errors are especially undesirable.
Where yi is the actual value, ŷi is the predicted value, and n signifies
the total number of observations
∑(yi −ŷi )2
R2 (Coefficient of De- This statistical metric reflects the proportion of variance in the depen- R2 = 1 −
∑(yi −ȳ)2
termination) dent variable that can be attributed to the independent variables in a
regression analysis. It evaluates the adequacy of the model, demon-
strating the extent to which the predicted values correspond with
the actual data. A perfect R2 of 1 means the predictions match the
data exactly.
Where yi is the actual value, ŷi is the predicted value, ȳ is the mean
of actual values, and n is the number of observations.

For a comprehensive evaluation of the model’s performance, the inclusion of MAE,

MSE, and RMSE as complementary metrics is justified by:
- MAE provides an average measure of absolute error.
- MSE emphasizes more significant errors by penalizing them quadratically.
- RMSE offers an interpretable error measure in the same units as the target variable,
allowing us to analyze the models from different perspectives.

4. Experimental Results
In this section, we estimate the EV battery SOC based on data-driven approaches by
considering impact factors that can significantly influence energy consumption. We use
two different datasets [47,48] under two different contexts:
• Dataset 1: Referred to as Munich20, it consists of 72 actual driving recorded trips in
Munich, Germany, of a BMW I3 (60 Ah), highlighting the HVAC system’s significant
power demand. This dataset validates a comprehensive vehicle model, including
the powertrain and the HVAC. Each trip contains environmental, vehicle, battery,
and heating circuit data. It includes Category A (recorded trips in summer) and
Category B (winter).
J. Low Power Electron. Appl. 2024, 14, 59 9 of 16

• Dataset 2: known as HELECAR-D, is a dataset collected in Morocco from three routes

within the cities of Rabat and Sala Al Jadida. The data were gathered using a Renault
TWIZY vehicle and includes various features such as date, time, battery SOC, vehicle
speed, driving mode, GPS data, weather conditions, traffic information, and speed
limits. It is worth noting that two recorded trajectories took place on urban roads,
and the dataset includes 3 categories: T1 (light traffic), T2 (dense traffic), and T3
(moderate traffic). The works of [27] have used this dataset for SOC forecasting using
a hybrid approach as stated in the related work section.
We reconstituted the databases for the second test evaluation as follows:
1. Munich20 Database:
(a) DB_test1_Munich20: A mix of trips A (70%) and B (30%). It is divided into 2/3
for training and 1/3 for testing.
(b) DB_test2_Munich20: A mix of trips A (30%) and B (70%), used entirely for
testing (100%).
2. HELECAR-D Database:
(a) DB_test1_HELECAR-D: This dataset includes a mix of trips with light traffic
(T1, 50%) and dense traffic (T2, 50%). It is divided into 2/3 for training and
1/3 for testing.
(b) DB_test2_HELECAR-D: This dataset comprises trips with moderate traffic (T3)
and is used entirely for testing (100%).
Data preprocessing consists of handling missing values and outliers of each trip
and then integrating values of chosen trips to have a large dataset for better model training;
the input features are then selected as presented in Figure 3.

(a)

(b)

Figure 3. NN architecture with selected features for each driving scenario ((a) Munich20,
(b) HELECAR-D 2).
J. Low Power Electron. Appl. 2024, 14, 59 10 of 16

Finally, all obtained data are normalized to mitigate fluctuations associated with the
varying scales of each feature. Both datasets are split into training and testing as highlighted
previously. All the chosen algorithms were tested twice on the corresponding test sets, and
each algorithm’s performance should be assessed using error metrics as indicated in the
flow chart of the proposed evaluation methodology.
In this study, we developed two separate neural network (NN) models for SOC
estimation to account for the differences in features between the datasets. The MUNICH20
dataset includes HVAC energy consumption as a feature but does not provide information
on driving modes, whereas the HELECAR-D dataset lacks HVAC energy consumption but
includes driving modes. These features are critical as they are strongly correlated with SOC
changes. Training a single NN model across both datasets would introduce significant bias.
For instance, including the driving mode feature in the MUNICH20 model would always
result in the feature’s value being zero, which could skew the model’s learning process.
Similarly, adding HVAC energy demand to the HELECAR-D model would misrepresent
the data since the vehicle used in this dataset does not have an HVAC system. To ensure
accurate learning from the available features in each dataset, two independent NN models
were trained, each optimized for its respective dataset.
First, we trained regression models (LR, SVR, and RF) on both datasets.
Figures 4 and 5 present the estimated SOC output.

Figure 4. Recorded SOC vs. Estimated SOCusing Munich20.

J. Low Power Electron. Appl. 2024, 14, 59 11 of 16

Figure 5. Recorded SOC vs. Estimated SOC using HELECAR-D.

(a) Regression performance

In the first scenario as well as the second, all the regression algorithms have per-
formed commendably in estimating the SOC values over time on the first test sets.
The RF algorithm showed a relatively lower performance compared to LR and SVR
for Munich20, with an MSE of 3.44 and R² of 0.972, but was still fairly accurate.
For HELECAR-D, RF excelled in the first test with the lowest MSE of 1.20 and the
highest R² of 0.997, indicating very accurate estimations. However, to validate the
regression models for SOC forecasting, we must consider the different input fea-
tures selected in both datasets. It is prudent to test on a new set for both scenarios.
Table 2 presents the performance indices (PI) of the regression models on both tests
for each dataset. Notably, the second test represents actual trips considered only for
a second-test model evaluation. Performance of all algorithms decreased in Test 2
compared to Test 1, with SVR maintaining the best balance of error metrics (MSE of
2.32 and MAE of 1.27) and R² (0.73), indicating relatively better predictive capability
for Munich20. However, the performance on the second test set for both datasets
decreases, evidenced by LR’s R² dropping to 0.56 in Munich20, Test 2, even after
extensive hyperparameter tuning and model regularization. Most regression models
show promise when splitting the datasets into training, validation, and test sets due
J. Low Power Electron. Appl. 2024, 14, 59 12 of 16

to the model’s tendency to overfit with large datasets and numerous features, even
after thorough data preprocessing. Consequently, it is essential to conduct additional
testing on different scenarios before validating the models.
(b) Neural Network performance
The NN training takes longer time than the previous algorithms; however, it has also
performed well in the first test sets for both scenarios, as presented in Figures 6 and 7
and also in the second test sets, as in Table 2, compared to the ML performance on the
second test set. These findings demonstrate that the modeled NN for both datasets
can handle overfitting much better than the ML algorithms.

Figure 6. Recorded SOC vs. NN-Estimated SOC using Munich20.

Figure 7. Recorded SOC vs. NN-Estimated SOC using HELECAR-D.

Across most tests, neural networks (NN) consistently outperformed other algorithms,
especially in Munich20, where they exhibited remarkable adaptability to various data
conditions, reflected in their low error rates (MSE of 0.62) and high R² value (0.992) in
the first test. Conversely, random forest’s (RF) performance varied: it was outstanding
in the first test of HELECAR-D (with an MSE of 1.20 and R² of 0.997) but less impressive
in other scenarios, particularly in the second test of Munich20, where it had an MSE of
7.29 and an R² of 0.75. While reliable, linear regression (LR) and support vector regression
(SVR) did not reach the performance levels of NN and were occasionally surpassed by RF,
as seen with LR’s R² of 0.56 in Munich20, Test 2. This fluctuation in effectiveness across
different tests and datasets underscores the impact of specific data characteristics on the
J. Low Power Electron. Appl. 2024, 14, 59 13 of 16

algorithms’ performance. NN’s strong performance is likely due to their superior capability
in modeling complex and non-linear relationships, a feature less effectively captured by
the other models. Although each algorithm has its advantages, NN emerged as the most
flexible and effective across the conducted tests and datasets, as shown by their consistent
R² values above 0.9 in all but one scenario. Specifically, in the second test of Munich20, NN
achieved an MSE of 0.62 and an R² of 0.92, and in HELECAR-D, Test 2, they maintained
a high R² of 0.96 despite the challenges. These findings highlight NN’s robustness and
adaptability in various testing conditions.

Table 2. Comparison of algorithm performance using Munich20 and HELECAR-D.

Dataset Test Algorithm MSE MAE RMSE R²

Munich20 1 LR 2.82 1.29 1.68 0.97
Munich20 1 SVR 2.92 1.27 1.71 0.976
Munich20 1 RF 3.44 1.31 1.85 0.972
Munich20 1 NN 0.62 0.49 0.79 0.995
Munich20 2 LR 3.83 1.71 1.96 0.56
Munich20 2 SVR 2.32 1.27 1.52 0.73
Munich20 2 RF 7.78 2.21 2.79 0.11
Munich20 2 NN 0.62 0.65 0.79 0.92
HELECAR-D 1 LR 2.29 1.26 1.51 0.994
HELECAR-D 1 SVR 2.63 1.24 1.62 0.993
HELECAR-D 1 RF 1.20 0.93 1.10 0.997
HELECAR-D 2 NN 3.60 1.29 1.90 0.96
HELECAR-D 2 LR 37.84 5.64 6.15 0.65
HELECAR-D 2 SVR 17.25 3.55 4.15 0.84
HELECAR-D 2 RF 70.39 6.80 8.39 0.36
HELECAR-D 1 NN 3.63 1.51 1.91 0.991

In electric mobility research, accurately estimating the SOC is critical for operational
reliability and efficiency. This work contributes to the field by using real-world driving data,
contrasting with most existing literature, which uses battery data from controlled laboratory
conditions as presented in the works of [9,24,32] or overly simplified driving scenarios.
Our findings indicate that real-world data, including a more extensive range of driving
scenarios and auxiliary components like HVAC systems, weather conditions, and specific
route information, results in a more significant difference between Estimated SOC and
Recorded SOC values from the datasets. This deviation emphasizes the complexities
of real-world EV operation, as opposed to the idealized settings commonly depicted in
simulation-based research. In our analysis, neural networks (NN) demonstrated superior
performance in handling these complexities, especially when compared to more traditional
models like linear regression (LR) and support vector regression (SVR). The NN’s ability to
minimize MSE and maximize R² is particularly notable in HELECAR-D. The findings align
with results from [27] that employed a hybrid model using LR and LSTM networks, which
also leveraged real-world data. However, our research adds more to the conversation
by integrating an additional test set and additional datasets that highlight the energy
consumption of HVAC systems, a critical factor that should not be overlooked in SOC
estimation for BEVs. Moreover, our research incorporated an essential second testing phase,
a step often neglected in related studies. This second phase is crucial for reducing the
risk of overfitting, which happens when models are too fine-tuned to a particular dataset,
making them less effective in real-world situations. This part of our study helps ensure
J. Low Power Electron. Appl. 2024, 14, 59 14 of 16

that the utilized NN models are robust and can be trusted to work well across different
driving conditions.

5. Conclusions
Battery state of charge (SOC) estimation in electric vehicles is crucial for intelligent
energy management; therefore, in this paper, we evaluated four different algorithms: linear
regression (LR), support vector regression (SVR), random forest regression (RF), and neural
networks (NN) on two datasets from different contexts (European and African). With each
specific number of features for SOC estimation, two tests were applied on both algorithms
for each dataset. It can be concluded based on the findings that NNs consistently displayed
high predictive accuracy across different scenarios in Munich20, surpassing models with the
lowest MSE and highest R² values. While LR and SVR demonstrated commendable model
fits with high R² values, NN generally outperformed them. RF exhibited a more variable
performance, with relatively lower accuracy than LR and SVR in Munich20, but excelled
in HELECAR-D, Test 1, with the best predictive results among the algorithms. However,
RF’s performance notably declined in other tests. SVR emerged as a robust algorithm in
HELECAR-D, Test 2, showing a favorable balance between error metrics and R², suggesting
a reliable performance even when other algorithms faltered. Overall, while each algorithm
had strengths in specific tests, NN proved to be the most consistent regarding predictive
accuracy. RF showed the ability to achieve high accuracy under certain conditions.
In upcoming projects, we plan to use the neural network (NN) models to estimate the
state of charge (SOC) for a vehicle’s journey. This estimation will then make intelligent
decisions about distributing the vehicle’s energy. By knowing the SOC in advance, we can
better plan how to use the energy throughout the trip. This could help in using less power
when it is not needed, which saves energy, or ensuring there is enough power for the parts
of the trip that require more, like uphill stretches. The goal is to manage the vehicle’s energy
most effectively, making the ride smoother and the vehicle’s energy use more efficient.

Author Contributions: S.A.: Conceptualization, Methodology, Data curation and pre-processing,

Algorithm testing and validation, Writing—original draft. Z.B., A.E.O. and A.E.: Data curation,
Formal analysis, Visualization, Writing—review & editing. A.T. and S.E.: Supervision, Project
administration, Reviewing and Editing, Funding acquisition. All authors have read and agreed to the
published version of the manuscript.
Funding: This project has received funding from: The Ministry of Europe and Foreign Affairs (MEAE),
the Ministry of Higher Education and Research (MESR), and the Ministry of Higher Education,
Scientific Research, and Innovation (MESRSI). The funding falls under the Franco-Moroccan Hubert
Curien Partnership (PHC-TOUBKAL) program 2022–2023, with Grant number: 48530UC.
Data Availability Statement: We have provided the link to the datasets used in this paper [47,48].
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. López-Claros, A.; Dahl, A.L.; Groff, M. Global Governance and the Emergence of Global Institutions for the 21st Century; Cambridge
University Press: Cambridge, UK, 2020. [CrossRef]
2. Hossain; Kumar, L.; Assad, M.E.H.; Alayi, R. Advancements and Future Prospects of Electric Vehicle Technologies: A Compre-
hensive review. Complexity 2022, 2022, 3304796. [CrossRef]
3. International Energy Agency (IEA). Global EV Outlook 2021. 2021. Available online: https://www.iea.org/reports/global-ev-
outlook-2021 (accessed on 15 January 2024).
4. She, Z.Y.; Sun, Q.; Ma, J.J.; Xie, B.C. What are the barriers to widespread adoption of battery electric vehicles? A survey of public
perception in Tianjin, China. Transp. Policy 2017, 56, 29–40. [CrossRef]
5. Lu, L.; Han, X.; Li, J.; Hua, J.; Ouyang, M. A review on the key issues for lithium-ion battery management in electric vehicles. J.
Power Sources 2013, 226, 272–288. [CrossRef]
6. Zhang, M.; Fan, X. Review on the state of charge estimation methods for electric vehicle battery. World Electr. Veh. J. 2020, 11, 23.
[CrossRef]
J. Low Power Electron. Appl. 2024, 14, 59 15 of 16

7. Bhatt, D.K.; Darieby, M.E. An Assessment of Batteries form Battery Electric Vehicle Perspectives. In Proceedings of the 2018 IEEE
International Conference on Smart Energy Grid Engineering (SEGE), Oshawa, ON, Canada, 12–15 August 2018; pp. 255–259.
[CrossRef]
8. Norouzi, A.; Heidarifar, H.; Borhan, H.; Shahbakhti, M.; Koch, C.R. Integrating Machine Learning and Model Predictive Control
for automotive applications: A review and future directions. Eng. Appl. Artif. Intell. 2023, 120, 105878. [CrossRef]
9. Zhao, J.; Ling, H.; Liu, J.; Wang, J.; Burke, A.; Lian, Y. Machine learning for predicting battery capacity for electric vehicles.
eTransportation 2023, 15, 100214. [CrossRef]
10. Zekić–Sušac, M.; Mitrović, S.; Has, A. Machine learning based system for managing energy efficiency of public sector as an
approach towards smart cities. Int. J. Inf. Manag. 2021, 58, 102074. [CrossRef]
11. Musbah, H.; Aly, H.H.H.; Little, T.A. Energy management of hybrid energy system sources based on machine learning
classification algorithms. Electr. Power Syst. Res. 2021, 199, 107436. [CrossRef]
12. Singirikonda, S.; Obulesu, Y.P. Battery modelling and state of charge estimation methods for Energy Management in Electric
Vehicle—A review. IOP Conf. Ser. 2020, 937, 012046. [CrossRef]
13. Naguib, M.; Kollmeyer, P.J.; Emadi, A. Lithium-Ion battery pack robust state of charge estimation, cell inconsistency, and
balancing: Review. IEEE Access 2021, 9, 50570–50582. [CrossRef]
14. Zhou, W.; Zheng, Y.; Pan, Z.; Lu, Q. Review on the battery model and SOC estimation method. Processes 2021, 9, 1685. [CrossRef]
15. Hannan, M.A.; Lipu, M.S.H.; Hussain, A.; Mohamed, A. A review of lithium-ion battery state of charge estimation and
management system in electric vehicle applications: Challenges and recommendations. Renew. Sustain. Energy Rev. 2017,
78, 834–854. [CrossRef]
16. Wang, Y.; Tian, J.; Sun, Z.; Wang, L.; Xu, R.; Li, M.; Chen, Z. A comprehensive review of battery modeling and state estimation
approaches for advanced battery management systems. Renew. Sustain. Energy Rev. 2020, 131, 110015. [CrossRef]
17. Wu, Y.; Sundaresan, S.; Balasingam, B. Battery Parameter Analysis through Electrochemical Impedance Spectroscopy at Different
State of Charge Levels. J. Low Power Electron. Appl. 2023, 13, 29. [CrossRef]
18. Cheng, K.W.E.; Divakar, B.; Wu, H.; Ding, K.; Ho, H.F. Battery-Management System (BMS) and SOC development for electrical
vehicles. IEEE Trans. Veh. Technol. 2011, 60, 76–88. [CrossRef]
19. Plett, G.L. Sigma-point Kalman filtering for battery management systems of LiPB-based HEV battery packs. J. Power Sources 2006,
161, 1356–1368. [CrossRef]
20. Zhao, Y.; Xie, W.; Wu, J. A novel SOC estimation method for supercapacitor cell module based on EKF-MF hybrid filtering
algorithm. J. Electr. Eng. Technol. Electr. Eng. Technol. 2024, 19, 4927–4940. [CrossRef]
21. Sepasi, S.; Ghorbani, R.; Liaw, B.Y. SOC estimation for aged lithium-ion batteries using model adaptive extended Kalman filter.
In Proceedings of the 2013 IEEE Transportation Electrification Conference and Expo (ITEC), Detroit, MI, USA, 16–19 June 2013;
pp. 1–6. [CrossRef]
22. How, D.N.; Hannan, M.; Lipu, M.H.; Ker, P.J. State of charge estimation for lithium-ion batteries using model-based and
data-driven methods: A review. IEEE Access 2019, 7, 136116–136136. [CrossRef]
23. Castanho, D.S.; Guerreiro, M.; Silva, L.C.A.; Eckert, J.J.; Alves, T.A.; De Souza Tadano, Y.; Stevan, S.L.; Siqueira, H.V.; Corrêa,
F.C. Method for SOC estimation in Lithium-Ion batteries based on multiple linear regression and particle swarm optimization.
Energies 2022, 15, 6881. [CrossRef]
24. Jumah, S.; Elezab, A.; Zayed, O.; Ahmed, R.; Narimani, M.; Emadi, A. State of Charge Estimation for EV Batteries Using Support
Vector Regression. In Proceedings of the 2022 IEEE Transportation Electrification Conference & Expo (ITEC), Anaheim, CA, USA,
15–17 June 2022; pp. 964–969. [CrossRef]
25. Venkatesan, C.; Patil, C.K.; Karthick, A.; Dharmaraj, G.; Rahim, R.; Ghosh, A. State of charge estimation of Lithium-Ion battery
for electric vehicles using machine learning algorithms. World Electr. Veh. J. 2021, 12, 38. [CrossRef]
26. Li, C.; Chen, Z.; Cui, J.; Wang, Y.; Zou, F. The lithium-ion battery state-of-charge estimation using random forest regression. In
Proceedings of the 2014 Prognostics and System Health Management Conference (PHM-2014 Hunan), Zhangjiajie, China, 24–27
August 2014; pp. 336–339. [CrossRef]
27. NaitMalek, Y.; Najib, M.; Lahlou, A.; Bakhouya, M.; Gaber, J.; Essaaidi, M. A hybrid approach for State-of-Charge forecasting in
Battery-Powered electric vehicles. Sustainability 2022, 14, 9993. [CrossRef]
28. Wang, Q.; Sun, C.; Gu, Y. Research on SOC estimation method of hybrid electric vehicles battery based on the grey wolf optimized
particle filter. Comput. Electr. Eng. 2023, 110, 108907. [CrossRef]
29. Das, K.; Kumar, R. A comprehensive review of categorization and perspectives on State-of-Charge estimation using deep learning
methods for electric transportation. Wirel. Pers. Commun. 2024, 133, 1599–1618. [CrossRef]
30. Lee, D.T.; Shiah, S.J.; Lee, C.M.; Wang, Y.C. State-of-charge estimation for electric scooters by using learning mechanisms. IEEE
Trans. Veh. Technol. 2007, 56, 544–556. [CrossRef]
31. Guo, S.; Ma, L. A comparative study of different deep learning algorithms for lithium-ion batteries on state-of-charge estimation.
Energy 2023, 263, 125872. [CrossRef]
32. El Fallah, S.; Kharbach, J.; Hammouch, Z.; Rezzouk, A.; Jamil, M.O. State of charge estimation of an electric vehicle’s battery
using Deep Neural Networks: Simulation and experimental results. J. Energy Storage 2023, 62, 106904. [CrossRef]
33. Chemali, E.; Kollmeyer, P.J.; Preindl, M.; Emadi, A. State-of-charge estimation of Li-ion batteries using deep neural networks: A
machine learning approach. J. Power Sources 2018, 400, 242–255. [CrossRef]
J. Low Power Electron. Appl. 2024, 14, 59 16 of 16

34. Azkue, M.; Miguel, E.; Martinez-Laserna, E.; Oca, L.; Iraola, U. Creating a Robust SoC Estimation Algorithm Based on LSTM
Units and Trained with Synthetic Data. World Electr. Veh. J. 2023, 14, 197. [CrossRef]
35. Bockrath, S.; Rosskopf, A.; Koffel, S.; Waldhör, S.; Srivastava, K.; Lorentz, V.R. State of charge estimation using recurrent neural
networks with long short-term memory for lithium-ion batteries. In Proceedings of the IECON 2019-45th Annual Conference of
the IEEE Industrial Electronics Society, Lisbon, Portugal, 14–17 October 2019; Volume 1, pp. 2507–2511. [CrossRef]
36. Li, X.; Jiang, H.; Guo, S.; Xu, J.; Li, M.; Liu, X.; Zhang, X. SOC estimation of Lithium-Ion battery for electric vehicle based on deep
multilayer perceptron. Comput. Intell. Neurosci. 2022, 2022, 3920317. [CrossRef]
37. Kim, S.; Lim, H. Reinforcement learning based energy management algorithm for smart energy buildings. Energies 2018, 11, 2010.
[CrossRef]
38. NaitMalek, Y.; Najib, M.; Bakhouya, M.; Essaaidi, M. Embedded real-time battery State-of-Charge forecasting in Micro-Grid
systems. Ecol. Complex. 2021, 45, 100903. [CrossRef]
39. Ariche, S.; Boulghasoul, Z.; El Ouardi, A.; Elbacha, A.; Tajer, A.; Espie, S. Energy Consumption of a Battery Electric Vehicle for a
Comfort Ride on Moroccan Roads. In Proceedings of the 2023 9th International Conference on Control, Decision and Information
Technologies (CoDIT), Rome, Italy, 3–6 July 2023; pp. 1–5. [CrossRef]
40. Electric Vehicle Thermal Management. Mathworks. Available online: https://www.mathworks.com/help/hydro/ug/sscfluids_
ev_thermal_management.html (accessed on 25 November 2024).
41. Vehicle Electrical and Climate Control Systems. Mathworks. Available online: https://www.mathworks.com/help/simulink/
slref/vehicle-electrical-and-climate-control-systems.html (accessed on 25 November 2024).
42. EV Battery Cooling System. Mathworks. Available online: https://www.mathworks.com/help/hydro/ug/ev-battery-cooling.
html (accessed on 25 November 2024).
43. Pedregosa, F.; Varoquaux, G.; Gramfort, A.; Michel, V.; Thirion, B.; Grisel, O.; Blondel, M.; Prettenhofer, P.; Weiss, R.; Dubourg, V.;
et al. Scikit-learn: Machine learning in Python. J. Mach. Learn. Res. 2011, 12, 2825–2830.
44. Abadi, M.; Agarwal, A.; Barham, P.; Brevdo, E.; Chen, Z.; Citro, C.; Corrado, G.S.; Davis, A.; Dean, J.; Devin, M.; et al. TensorFlow:
Large-Scale Machine Learning on Heterogeneous Distributed Systems. arXiv 2016, arXiv:1603.04467.
45. Chollet, F.; et al. Keras. 2015. Available online: https://keras.io (accessed on 15 April 2024).
46. Naser, M.; Alavi, A.H. Error metrics and performance fitness indicators for artificial intelligence and machine learning in
engineering and sciences. Archit. Struct. Constr. 2021, 3, 499–517. [CrossRef]
47. Steinstraeter, M. Battery and Heating Data in Real Driving Cycles. IEEE Dataport. 2020. Available online: https://ieee-dataport.
org/open-access/battery-and-heating-data-real-driving-cycles (accessed on 24 November 2024).
48. NaitMalek, Y.; Najib, M.; Bakhouya, M.; Gaber, J. HELECAR-D: A dataset for urban electro mobility in Moroccan context. Data
Brief 2023, 48, 109080. [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.