See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/361492509

Battery optimization by machine learning algorithms: Research gap via

bibliometric analysis

Conference Paper · June 2022

CITATIONS READS

0 935

11 authors, including:

Nico Hananda Azure Kamul

Binus University Binus University
8 PUBLICATIONS 19 CITATIONS 4 PUBLICATIONS 1 CITATION

SEE PROFILE SEE PROFILE

Christian Harito Endang Djuana

Binus University Universitas Trisakti
62 PUBLICATIONS 489 CITATIONS 35 PUBLICATIONS 84 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Polymer nanocomposites View project

The Investigation of the Effectiveness of Blended Learning View project

All content following this page was uploaded by Bens Pardamean on 17 May 2023.

The user has requested enhancement of the downloaded file.

E3S Web of Conferences 388, 01020 (2023) https://doi.org/10.1051/e3sconf/202338801020
ICOBAR 2022

Battery optimization by machine learning algorithms: Research

gap via bibliometric analysis
Nico Hananda1,*, Azure Kamul1, Christian Harito1, Endang Djuana2, Gregorius Natanael Elwirehardja3,4,
Bens Pardamean4, Fergyanto E. Gunawan5, Arief S. Budiman5, Muhammad Asrol5, A. A. N. Perwira Redi5, Tim Pasang6
1Industrial Engineering Department, Faculty of Engineering, Bina Nusantara University, Jakarta, 11480, Indonesia
2Electrical Engineering Department, Faculty of Industrial Technology, Universitas Trisakti, Jakarta, 11440, Indonesia
3 Bioinformatics and Data Science Research Center, Bina Nusantara University, Jakarta, Indonesia, 11480
4 Computer Science Department, BINUS Graduate Program - Master of Computer Science Program, Bina Nusantara University,

Jakarta, Indonesia, 11480

5Engineering Industrial Engineering Department, BINUS Graduate Program - Master of Industrial Engineering, Bina Nusantara

University, Jakarta, Indonesia 11480

6Department of Mechanical and Manufacturing Engineering and Technology, Oregon Institute of Technology, Klamath Falls, OR

97601, USA

Abstract. Technological developments enable low-carbon transitions to be accelerated by conceptualization systems and
innovations for research and development to generate clean energy. Batteries are becoming one of the essential parts of
the science of electrical power sources. Lithium-ion batteries are part of the change and development factors in
technologies that significantly impact the portable devices sector and the development of electric vehicles. Designing the
material structure and composition of battery manufacturing with the help of engineering system design will form a much
more optimal battery. Machine learning algorithms can easily optimize the battery’s composition through battery
experiment test data history to produce a more optimal battery configuration. This study is prepared to identify research
gaps in topics related to machine learning for battery optimization. Related studies about machine learning for battery
optimization are identified using bibliometric analysis and systematic literature review of the study search index through
database Scopus-indexed publications. The results from this paper reveal energy management systems and strategies,
hybrid vehicles, other optimization algorithms, battery electrodes, and the safety of batteries as the particular research
gap according to machine learning for battery optimization. This paper expects research on battery optimization using
machine learning methods will continue to be developed to maximize the potential of machine learning algorithms in
helping the research process.

batteries are part of the change and development factors

in technologies that significantly impact the portable
1. Introduction devices sector and the development of electric vehicles
Technological developments are enabling low- [2]. Optimization of the battery is needed to increase
carbon transitions to be accelerated and have made the battery life and increase the power flow channeled
idea of clean energy disruption widespread. The through the battery. Various methods have been
transition to clean energy aims to reduce global CO2 developed to accelerate the battery optimization process
emissions. Through the clean energy transition, it will in manufacturing, industrial, and electronic needs.
have an impact on around 40% of global CO2 emissions Given that energy needs are increasing, the construction
[1]. Conceptualization systems and innovations are able of optimal battery design is increasingly being
to accelerate research and development to generate developed using constantly evolving technology. [3].
clean energy. Electrical energy is one of the clean The method of the battery optimization structure is
energies free from carbon emissions and easily obtained carried out to get a battery formulation with high power
today. This is due to electrical power sourced from resistance and energy distribution for the needs of the
renewable energy and does not require a combustion technology transition era. Designing the material
process that causes an increase in carbon emissions. The structure and composition of battery manufacturing with
source of electrical energy obtained through various the help of engineering system design will form a much
renewable energy is channeled and accommodated more optimal battery. The purpose of the invention is to
through batteries. maximize the cycle in the battery to be more reliable
through charging data and power distribution in the
*Corresponding author: [email protected] battery. Data generated from battery life trials will be
Batteries are becoming one of the essential parts of the used as parameters translated by machine learning to
science of electrical power sources. Lithium-ion create more optimal design batteries, have a longer cycle

© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0
(https://creativecommons.org/licenses/by/4.0/).
E3S Web of Conferences 388, 01020 (2023) https://doi.org/10.1051/e3sconf/202338801020
ICOBAR 2022

life, deliver a high power, and are relatively affordable optimizing battery storage systems for battery
using environmentally friendly materials [4]. manufacturing optimization design. Different parameter
There are millions of possibilities found suitable possibilities should be considered simultaneously to
materials to be used as electrolytes in batteries. Machine create an optimized battery storage system [6]. The
learning technology is able to help filter optimal ability of machine learning depends mainly on the
materials for battery development. Machine learning is quantity, quality, and veracity of the data. This makes
part of a computer system related to artificial the research and development process necessary to
intelligence stands out as a promising approach for prepare enough data sets to expand simulations and
research and development, especially in battery machine thinking knowledge during the machine
optimization. Artificial intelligence and machine learning process [8]. Afterward, the architecture in
learning are able to solve problem parameters machine learning algorithms needs to be trained through
effectively for a more optimal battery technologies simulations and experiments to create supervised
research and development process [5]. For example, models in machine learning models. The training results
when developing electrochemical capacitors and conducted through simulations will continue to be
electrolyte fluids in batteries related to experimental recorded and compared to produce outputs in the form
data. Machine learning algorithms can easily optimize of machine learning algorithms which will then be used
the battery’s composition through battery experiment for design parameters in the battery optimization
test data history to produce a more optimal battery research and development process.
configuration. In addition, machine learning is able to
expand the scale of the approach to get a more optimal
battery design. Accurate data processes and able to
identify errors make machine learning able to simulate
experiments on a larger scale without experiments that
require high costs [6]. Transparent data access makes
the opportunity to adopt machine learning systems for
research and development toward battery optimization
even more promising. Science and technological
developments make machine learning systems capable Fig.1. Working Principle of Machine Learning Algorithm
of processing data with a high level of accuracy. It is
helpful for the study and future development of battery
optimization [7]. Machine learning algorithms can be classified into three
This study is prepared to identify research gaps in topics levels of understanding: unsupervised, semi-supervised,
related to machine learning for battery optimization. and supervised (see Figure 1) [9] [10]. According to
Related studies about machine learning for battery unsupervised, the machine learning approach uses raw
optimization are identified using bibliometric analysis data processed by machine learning. This approach aims
and systematic literature review of the study search to design data patterns that will be processed to form
index through database Scopus-indexed publications. machine learning algorithms. The unsupervised
During the identification of various related studies, machine learning approach will provide data output
researchers use the Publish or Perish application as a which will then be rerun by machine learning called the
corresponding study search index combined with supervised machine learning approach [9]. Supervised
VOSviewers to display bibliometric analysis. This study machine learning will display the results of regression
aims to explain methods or algorithms that have a strong and classification used to describe and conclude data
relationship using bibliometric analysis to show the simulations carried out at the unsupervised machine
results of studies classified as high impact to help in the learning stage [11]. The output provided from
development process and further studies related to supervised machine learning is in algorithmic designs
battery optimization using machine learning. such as regression and classification as decision-making
parameters in the research and development process.
Meanwhile, semi-supervised approaches are a process
2. Literature Review that is located when data generated from the
unsupervised stage begins to produce data that can be
2.1. Data-driven machine learning approach for run at the supervised approaches stage. The multi-output
battery R&D method is proposed through data-driven models
resulting from machine learning to create optimal
Historically, machine learning can be defined as how battery formulation in designing battery optimization
a machine is able to improve its performance as a [12].
machine to achieve specific goals through data and
simulations. The ability to automatically conduct
experimental processes from machine learning systems
2.2. Battery Energy Storage System
is one of the approaches to improving research and
Optimization
development performance in battery manufacturing
optimization [3]. Machine learning methods are Clean energy sources use renewable resources and
particularly relevant for material discovery or batteries as storage systems into an innovative and

2
E3S Web of Conferences 388, 01020 (2023) https://doi.org/10.1051/e3sconf/202338801020
ICOBAR 2022

environmentally friendly solution to be implemented as system, and environmental (see Figure 2) [14]. Charging
sustainable energy for future energy needs closely and discharging are constraints that refer to optimizing
related to technological developments. A battery energy maximum limits and minimum battery cycles, intending
storage system is a system that supports the continuity to extend the battery’s calendric cycle. Capacity
of the transition to the use of renewable resources as a constraint refers to reducing battery degradation due to
step to realizing clean and emission-free energy. The the charging-discharging process and regulating energy
most common battery energy technology today is distribution at a specific capacity to prevent damage to
lithium-ion batteries. The most common type of lithium the battery cell. System reliability constraints guide
battery found is LiMnO2 which is used for electronic controlling the battery system's reliability that needs to
devices [13]. The other commonly used lithium battery be ensured to limit overall power and energy to avoid
technology, LiNiMnCoO2 and Li2Mn2O4, are mainly damage to battery cells or life-shortening. Battery
used for medical devices. The main advantages of system optimization can determine the operational level
lithium-ion batteries are that they are portable, have a of the battery energy storage system when there is a
high energy density, and have fast charging power [14]. direct change in power output during the charge-
Besides that, the main drawbacks of lithium-ion discharge process. Environmental constraints are related
batteries are limited energy capacity and manufacturing to the materials used in batteries. The materials need to
costs that tend to be expensive. Optimizing the battery be optimized to maximize the usefulness of the storage
energy storage system needs to be developed to find system battery and ensure to use of materials that are not
better battery system materials and composition. Battery harmful to the environment. Identification of all
energy storage systems optimization helps the flow of constraint systems in conducting battery optimization
energy with fast response time, high reliability, and low can be traced through simulation and data analysis.
self-discharge rate to be a promising potential for Parameters and data constraints will be efficiently
developing battery storage systems. Optimizing the processed using a machine learning approach to solve
design of an efficient battery storage system is essential battery optimization problems [4]. Machine learning
to improve reliability and reduce net present cost (NPC) could be a promising approach to optimizing the battery
in battery manufacturing [15]. Battery optimization also storage system.
helps limit the emission of ozone-depleting substances
during the production process.
2.3. Bibliometric analysis & Study Search
Some factors and aspects considered as system
Engine Indexing
constraints in optimizing the battery storage system are
battery cell life, cost efficiency, charging and Bibliometric analysis is a popular method of
discharging operation, power oscillations, sudden load analyzing large amounts of scientific data. The
changes, disruption of the transmission system, and bibliometric analysis uses quantitative methods to
battery energy distribution [16]. The lifetime of a battery measure specific indicators in scientific data, studies,
depends on the cell structure used, energy channeling and papers. Bibliometric analysis techniques are divided
operations, and thermal environment battery placement into two types, namely performance analysis and
which is also affected by the charging-discharging cycle science mapping based on research papers [19] show the
[17]. Sometimes the life of a battery also depends on the goal of the bibliometric analysis is to summarize large
production cost. In addition, battery capacity and state amounts of bibliometric data to find emerging trends of
of health (SoH) are parameters for forming a battery research topics in structured network visualization. The
management system [1]. The battery management analysis technique performed using the bibliometric
system using data-driven methods has successfully method is used when a broad scope of research in a
projected the SoH of Li-ion batteries through particular field requires large scientific data sets. From
algorithms; it shows that the battery management the point of view of scientists and researchers, the
system using machine learning algorithm is able to do benefit of conducting bibliometric research is the
battery optimization such as improve battery quality, bibliometric method can easily classify and make the
reduce battery degradation for better capacity, and identification of the relationship between trending
increase the lifespan of the battery [18]. research topics with other research topics that have been
frequently researched, so the results of bibliometric
research can be an initial picture and study literature for
a research study in a particular field [20]. Whereas from
the point of view of the community and practitioners,
the benefits provided by bibliometric research can offer
indirect benefits to expand understanding of a particular
field. Thus, bibliometric study helps practitioners make
Fig.2. Optimization constraints for battery energy storage decisions from bibliometric results and networks [21].
system [14] Bibliometric methods are increasingly used, making
some application developers create software to perform
If inferred from various factors and considerations, bibliometric analysis [19]. The bibliometric tools used
battery energy storage system optimization has four in this study are VOSviewer as bibliometric network
constraints: charging and discharging, capacity, reliable visualization and Publish or Perish as search engine

3
E3S Web of Conferences 388, 01020 (2023) https://doi.org/10.1051/e3sconf/202338801020
ICOBAR 2022

indexing for citation analysis on academic metadata Table 1. Inclusion & Exclusion Criteria
from academic databases such as Scopus and Google Inclusion Paper Exclusion Paper
Scholar. Literature analysis from bibliometric analysis The scope of the analysis from scientific
Outside the scope of research on battery
combined with bibliometric mapping through network literature must be focused on machine learning
optimization study were excluded
for battery optimization
visualization with the VOSviewer application produces
an overview of the interrelationships between scientific Scientific literature must be journal article Review journal article were excluded

research using network approaches, areas (nodes), and Selected scientific literature must be published Articles that were published under 2015
from 2015 to 2022 were excluded
levels of interaction [22]. In addition, search engine Selected scientific literature must be written in
Other languanges were excluded
indexing from academic databases through the Publish English
or Perish application supports scientific data indexing as
a parameter to bibliometric mapping. Researchers can The detailed systematic literature searching process
see the relationship between keywords from scientific using PRISMA method is shown in Figure 3. In the first
data and sorted research about machine learning for step, when the keyword input "machine learning," there
battery optimization based on several criteria discussed were 424,158 studies that appeared. Then the search
in the methodology of this study. continued to be done by adding more keywords
"battery" and "optimization" to get related studies and
displayed 465 studies. The screening process of 465
3. Methodology study documents is filtered with the criteria in Table 1,
showing that there are 133 scientific articles after
3.1. Bibliometric Analysis & Mapping filtering by criteria process. The last screening is done
with the help of Publish or Perish and several stages of
This study's bibliometric analysis and mapping were manual identification to find scientific literature
performed through the Scopus database by searching identified with high impact to get 17 articles that will be
through search engine indexing available on the Scopus used for bibliometric analysis and systematic literature
database to find scientific literature from keywords review. Selected scientific literature studies through
about machine learning associated with battery several screening stages and software screening will be
optimization. Scientific literature searches are carried analyzed using VOSviewer software. The result will
out using Scopus's database because it has a significant show a visualization of associated keywords from
enough data completeness about books, journals, selected papers. The network of the related keywords
conference papers, thesis, abstracts, and citations. In will be used to identify research or publication gaps
addition, Scopus is also one of the largest digital referring to machine learning for battery optimization.
platforms that present a variety of science literacy and
has a user-friendly interface. The searching process for
relevant scientific literature data that will be taken as
parameters in bibliometric analysis and mapping will be
focused on narrowing the discussion scope through
several keywords. The elimination of study results that
are less relevant to the scope of discussions related to
machine learning for battery optimization is also carried
Fig.3. Scientific literature screening workflow with PRISMA
out to get more focused mapping and analysis results.
method
Keywords inputted on Scopus search engine indexing
include "Machine Learning”, “Battery”,
“Optimization”, and “Design” with input query on 4. Result and Discussion
Scopus database in the form of boolean logic written as
TITLE-ABS-KEY (machine AND learning AND
battery AND optimization AND algorithm). Publish or 4.1. Systematic literature review
Perish is used to find scientific literature related to Systematic literature reviews from 17 selected
machine learning for battery optimization that has a high papers using the Scopus database for the topic of
impact on providing several recommended studies about machine learning for battery management are shown in
machine learning for battery optimization. The results of Table 2. Through the 17 selected articles, it can be seen
recommendations and discussions from bibliometric that the use of machine learning algorithms in
analysis and mapping will be aimed for continuity of optimizing and helping systematic experiments on
studies and further research that discusses more battery optimization. Problems often encountered in the
comprehensively related to machine learning for battery battery optimization process are battery degradation,
optimization. Additional query was added to narrow the short life cycles, and malfunction in the internal part of
scope to more detailed and relevant based on inclusion the battery composition that causes heat in the battery
and exclusion criteria (see Table 1). Additional input in during the charging-discharging process. These battery
the form of boolean logic written as TITLE-ABS-KEY issues encourage battery optimization research to
(machine AND learning AND battery AND improve battery performance and optimal safety battery
optimization AND algorithm) AND (LIMIT-TO design through experiments and data processed through
(DOCTYPE, "ar") AND (LIMIT-TO (LANGUAGE, machine learning technology optimization algorithms
"English ")) AND PUBYEAR > 2015.

4
E3S Web of Conferences 388, 01020 (2023) https://doi.org/10.1051/e3sconf/202338801020
ICOBAR 2022

[23]. The algorithm approach used can be in several Type(s)

Approach
of Parameters Results Ref

forms of algorithmic techniques that have been Extreme learning

machine, bat
State
health
of The
model
ELM (Ge, Zhang, Kong, & Wan, 2022)

commonly used, such as various vector machine algorithm battery combined

with the bat
algorithms, genetic algorithms, particle swarm algorithm
creates faster
optimization, etc. But inventions related to novel learning and
high SOH
algorithms such as Forgetting Online Sequential estimation
accuracy for
Extreme Learning Machine (FOS-ELM) with the optimizing
the battery
Hybrid Grey Wolf Optimizer (HGWO) algorithm can be health
management
used for the Li-ion battery prognostic process and then Machine learning Optimal
system
There is a (Yamanaka et al., 2020)
can be used as a basis parameterization for optimizing framework,
genetic algorithm
lithium-ion
battery
high
correlation
battery design [24]. The results of the systematic design between fire
spread and
literature review also showed a link to data experiments negative
cathode.
and the use of algorithms to improve the accuracy of Using genetic
algorithms
experimental results and reduce costs during battery will help find
optimal
optimization, especially in battery cycle life coverage, design
conditions for
state of charge, state of health, energy management improved
design safety
system, battery capacity, electrochemical battery, and better
performance
optimizing battery electrode, and battery temperature Gaussian Naives The Machine (M Duquesnoy et al., 2021)
Bayes algorithm electrode in learning
[25]. Given that energy needs and the development of battery cell training
performance algorithms
electric vehicles are increasing, battery optimization and can predict
heterogeneity
the discovery of a novel mixture of electrochemical are and
homogeneity
increasingly essential to be studied further for the in electrodes
and become a
advancement of battery technology [26]. robust
optimization
method for
Table 2. Scope of discussion and methods related to battery next-
optimization from selected journal articles based on the generation
battery
search index of the Scopus database electrodes in
the
manufacturin
Type(s) of Parameters Results Ref g process
Approach Genetic Battery The particle (Yang, Peng, Wang, Yuan, & Wu, 2018)
Closed-Loop Battery Incorporates (Attia et al., 2020) algorithm, particle remaining swarm
Optimization cycle life, experimental swarm useful life algorithm is
(CLO), Predicting battery results to optimization (RUL) applied to
Final Cycles, material assist in algorithm improve the
Bayesian selection determining performance
Optimization future
Algorithm decisions and
generalized to
applications
in optimal 4.2. Bibliometric analysis & mapping
battery design
Improving back- State of BSA (Mahammad A. Hannan, Lipu, Hussain, Saad, & Ayob,
propagation neural charge, optimization 2018) The bibliometric method analyzes critical topics
network (BPNN) charging- utilization has
using backtracking discharging improved the related to machine learning, battery, and optimization.
search algorithm limit robustness of
(BSA) BPNN to This method will also help strengthen the search results
create a more
optimal state and analysis of the systematic literature review. Based
of charge
battery design on the 17 selected articles, several extracted keywords
Rule-based
algorithms,
Hybrid
energy
Machine
learning
(Xiong, Chen, Wang, & Sun, 2018)
from abstract and titles related to machine learning for
optimization-based
control algorithms,
storage
system
utilization has
improved the battery optimization are filtered and analyzed through
intelligent-based
control algorithms
optimization
(HESS) on
efficiency,
performance, VOSviewer and visualized in Figure
g 4.
battery and and lifetime
ultracapacit of HESS
or, energy through a
management multi-system
strategy control
strategy using
an energy
management
strategy
Deterministic Battery DDPG (Wu, Wei, Liu, Quan, & Li, 2020)
policy gradient temperature algorithm
algorithm (DDPG) and combined
with expert degradation with expert
assistance. control, assistance
battery creates a
health proposed
strategy that
can suppress
the battery
temperature
and reduce
overall
driving cost
compared to
existing
research Fig.4. Bibliometric analysis & mapping of machine learning
about
lithium-ion
the
related to battery optimization
battery
system

Table 3. Cont

5
E3S Web of Conferences 388, 01020 (2023) https://doi.org/10.1051/e3sconf/202338801020
ICOBAR 2022

to lithium-ion batteries and machine learning in Figure

4.

Fig.5. Density and overlay visualization of machine learning

related to battery optimization

Visualization results from the bibliometric mapping of

17 articles with critical topics in machine learning for
battery optimization are shown in Figure 4. Critical
issues related to machine learning, battery optimization,
and algorithms are divided into four types of clusters.
Cluster 1 is marked in red with 13 critical topics about
parameters closely associated with the battery
optimization process. Cluster 2 is marked in green with
eight critical topics related to approach algorithms
carried out during machine learning-related research and
experiment. Cluster 3 is marked in blue with seven Fig.6. The research gap related to machine learning,
critical topics related to energy management that closely optimization, and lithium-ion batteries
correlate with energy control and battery function.
Cluster 4 is observed in yellow with six crucial topics The bibliometric mapping results show a research gap
related to machine learning algorithm approaches and related to battery optimization through machine learning
optimization methods for lithium-ion battery algorithms. The research gap is about energy
optimization. It can be seen that cluster 4 has the most management systems and strategies, hybrid vehicles,
substantial link strength with another cluster. Cluster 4 other optimization algorithms, battery electrodes, and
contains critical topics such as machine learning and the safety of batteries (see Figure 6). Further research
lithium-ion batteries with solid relationship link scores related to energy management systems, hybrid vehicles,
above 50 link strengths compared to other essential electrochemical mixtures, and novel optimization
topics of the red and green clusters. The connections algorithms is essential to expand novel research related
show that algorithmic approaches, optimization to battery optimization using machine learning. The
methods, and parameters are related to lithium-ion limitations of this paper are searches that are carried out
batteries and the machine learning approach. In only using Scopus's database. The data extracted is only
addition, all kinds of optimization systems and in the form of keywords from titles and abstracts,
algorithms in conducting research methodologies have a bibliometric data, and author keywords.
strong relationship with machine learning for lithium-
ion battery optimization research and experiment. The
4.3. Future Recommendations
study focused on machine learning and optimization in
lithium-ion batteries is increasing. Research on Based on the systematic literature review and
optimization algorithms and machine learning is bibliometric analysis, there are several
expanding in 2020 as the world's energy needs increase recommendations for further studies about machine
(see Figure 5). The statement was supported by the learning for battery optimization. Considering the
emergence of the lithium-ion battery industry to fulfill development of technology and the increasing demand
the demand for electric vehicles to reduce world carbon for batteries, we need to expand and develop battery
emissions [4]. The strong link between optimization, technology, especially lithium-ion batteries that are
machine learning, and lithium-ion batteries suggests that commonly used. Problems related to temperature,
machine learning can increase accuracy in estimating battery performance, and battery optimization
experimental data for battery optimization and reduce constraints can still be explored more deeply in
research costs. This prompted many researchers to designing a more optimal and reliable battery. Then the
continue to develop new machine learning algorithms to electrochemical components in the battery can still be
expand the discovery of composition and design traced to get the right composition design to produce a
formulations on more optimal batteries in accordance more reliable battery with high performance. Further
with the discussion of optimization constraints in Figure explanation and research about the topics related to
2 and the network of optimization parameters connected research gap can be found in several papers from

6
E3S Web of Conferences 388, 01020 (2023) https://doi.org/10.1051/e3sconf/202338801020
ICOBAR 2022

previous section. The results of the literature review and will continue to be developed to maximize the potential
previous discussion also showed that several researchers of machine learning algorithms in helping the research
have optimized the battery using machine learning process. Machine learning can provide more accurate
algorithms with research parameters contained in the results and expand the discovery of algorithms and
bibliometric analysis, namely Teo Lombardo, Alejandro battery designs that are more optimal for future energy
A. Franco, Marc Duquesnoy, and Emiliano N. Primo needs.
(see Table 2). From the point of view of research and
experimentation using machine learning, the algorithms Acknowledgements
used in optimizing by machine learning can still be The research is funded by Penelitian International Binus
improved. Research and inventions related to algorithms (PIB) Bina Nusantara University under the code of
can still be done to maximize the potential of machine PIB08.
learning as an approach to conducting battery
optimization research by discovering novel optimization
algorithms. A deeper study on optimization using
machine learning can be started with research related to References
genetic algorithms or particle swarm optimization 1. P. Johnstone, K. S. Rogge, P. Kivimaa, C. F.
algorithms (see Table 2). Both optimization algorithms
Fratini, E. Primmer, and A. Stirling, Energy Res
can improve accuracy in conducting further research
related to batteries. In addition, analysis with new Soc Sci 59, 101287 (2020)
algorithms can also be developed to expand exploration 2. Y. Li, K. Liu, A. M. Foley, A. Zülke, M.
and maximize machine learning performance for future Berecibar, E. Nanini-Maury, J. Van Mierlo, and
research on battery optimization. Other research related H. E. Hoster, Renewable and Sustainable
to energy management systems, hybrid vehicles, Energy Reviews 113, 109254 (2019)
electrochemical mixtures, energy optimization cases, 3. S. X. Drakopoulos, A. Gholamipour-Shirazi, P.
and novel optimization algorithms will be vital literature MacDonald, R. C. Parini, C. D. Reynolds, D. L.
to close the research gap to further research related to Burnett, B. Pye, K. B. O’Regan, G. Wang, T. M.
machine learning for battery optimization. Whitehead, G. J. Conduit, A. Cazacu, and E.
Kendrick, Cell Rep Phys Sci 2, 100683 (2021)
5. Conclusion 4. T. Lombardo, M. Duquesnoy, H. El-Bouysidy,
F. Årén, A. Gallo-Bueno, P. B. Jørgensen, A.
Systematic Literature Review and Bibliometric Bhowmik, A. Demortière, E. Ayerbe, F.
Analysis are conducted in this paper. Searching process Alcaide, M. Reynaud, J. Carrasco, A. Grimaud,
for scientific literature is carried out by looking for C. Zhang, T. Vegge, P. Johansson, and A. A.
studies related to machine learning on the Scopus Franco, Chem Rev (2021)
database with the keywords “Machine Learning”, 5. A. Mistry, A. A. Franco, S. J. Cooper, S. A.
“Battery Optimization”, and “Algorithms”. From the
Roberts, and V. Viswanathan, ACS Energy Lett
424,158 studies about machine learning, it was
successfully filtered into 17 selected articles related to 6, 1422 (2021)
optimization, machine learning, and lithium-ion 6. T. Vegge, J. M. Tarascon, and K. Edström, Adv
batteries. Keyword data extracted from abstracts and Energy Mater 11, (2021)
titles of 17 selected papers and bibliometric data are 7. H. El-Bousiydy, T. Lombardo, E. N. Primo, M.
used as parameters for conducting bibliometric analysis. Duquesnoy, M. Morcrette, P. Johansson, P.
Systematic literature review results and bibliometric Simon, A. Grimaud, and A. A. Franco, Batter
analysis show a strong relationship between machine Supercaps 4, 758 (2021)
learning and lithium-ion battery optimization research. 8. C. Chen, Y. Zuo, W. Ye, X. Li, Z. Deng, and S.
The results of visualization through bibliometric P. Ong, Adv Energy Mater 10, (2020)
analysis show a research gap and strong relationships 9. S. Russell and R. Norvig, Artificial Intelligence:
related to machine learning for battery optimization as a A Modern Approach, 4th Edition, 4th ed.
foundation for developing research related to battery
(Pearson Education, 2020)
optimization that will affect the advancement of battery
10. T. Hastie, R. Tibshirani, and J. Friedman, The
technology. The results from this paper reveal energy
management systems and strategies, hybrid vehicles, Elements of Statistical Learning: Data Mining,
other optimization algorithms, battery electrodes, and Inference, and Prediction, Second Edition
the safety of batteries as the particular research gap (Springer Series in Statistics), Second
according to machine learning for battery optimization. (Springer, 2009)
In addition, several key papers and researchers who 11. M. Raissi, P. Perdikaris, and G. E. Karniadakis,
discuss machine learning algorithms for battery (2017)
optimization have been shown through systematic 12. S. Wang, S. Jin, D. Bai, Y. Fan, H. Shi, and C.
literature reviews and their relationship through Fernandez, Energy Reports 7, 5562 (2021)
bibliometric analysis. This paper expects research on
battery optimization using machine learning methods

7
E3S Web of Conferences 388, 01020 (2023) https://doi.org/10.1051/e3sconf/202338801020
ICOBAR 2022

13. H. Zheng, T. Wang, R. Zhao, J. Chen, and L. Li,

IOP Conf Ser Earth Environ Sci 108, 022011
(2018)
14. M. A. Hannan, S. B. Wali, P. J. Ker, M. S. A.
Rahman, M. Mansor, V. K.
Ramachandaramurthy, K. M. Muttaqi, T. M. I.
Mahlia, and Z. Y. Dong, J Energy Storage 42,
103023 (2021)
15. N. Ghorbani, A. Kasaeian, A. Toopshekan, L.
Bahrami, and A. Maghami, Energy 154, 581
(2018)
16. M. Baumann, M. Weil, J. F. Peters, N. Chibeles-
Martins, and A. B. Moniz, Renewable and
Sustainable Energy Reviews 107, 516 (2019)
17. L. Cai, J. Meng, D. I. Stroe, G. Luo, and R.
Teodorescu, J Power Sources 412, 615 (2019)
18. H. C. Hesse, M. Schimpe, D. Kucevic, and A.
Jossen, Energies 2017, Vol. 10, Page 2107 10,
2107 (2017)
19. N. Donthu, S. Kumar, D. Mukherjee, N. Pandey,
and W. M. Lim, J Bus Res 133, 285 (2021)
20. D. Mukherjee, W. M. Lim, S. Kumar, and N.
Donthu, J Bus Res 148, 101 (2022)
21. L. Radha and J. Arumugam, Shanlax
International Journal of Arts, Science and
Humanities 9, 44 (2021)
22. T. Saheb, B. Amini, and F. Kiaei Alamdari,
International Journal of Information
Management Data Insights 1, 100018 (2021)
23. T. Yamanaka, Y. Takagishi, and T. Yamaue, J
Electrochem Soc 167, 100516 (2020)
24. J. Fan, J. Fan, F. Liu, J. Qu, and R. Li, IEEE
Access 7, 160043 (2019)
25. P. M. Attia, A. Grover, N. Jin, K. A. Severson,
T. M. Markov, Y. H. Liao, M. H. Chen, B.
Cheong, N. Perkins, Z. Yang, P. K. Herring, M.
Aykol, S. J. Harris, R. D. Braatz, S. Ermon, and
W. C. Chueh, Nature 578, 397 (2020)
26. A. Dave, J. Mitchell, K. Kandasamy, H. Wang,
S. Burke, B. Paria, B. Póczos, J. Whitacre, and
V. Viswanathan, Cell Rep Phys Sci 1, (2020)

8
View publication stats