Advanced intelligent control for photovoltaic-vehicle-to-grid integration
Introduction. The increasing penetration of electric vehicles (EVs) and renewable energy has intensified concerns about grid stability and energy sustainability. Integrating photovoltaic (PV) systems with vehicle-to-grid (V2G) technology provides a promising solution but requires efficient energy ma...
Saved in:
| Date: | 2026 |
|---|---|
| Main Authors: | , , , |
| Format: | Article |
| Language: | English |
| Published: |
National Technical University "Kharkiv Polytechnic Institute" and Аnatolii Pidhornyi Institute of Power Machines and Systems of NAS of Ukraine
2026
|
| Subjects: | |
| Online Access: | https://eie.khpi.edu.ua/article/view/343893 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Journal Title: | Electrical Engineering & Electromechanics |
| Download file: |  |
Institution
Electrical Engineering & Electromechanics| _version_ | 1869562802979471360 |
|---|---|
| author | Lachheb, A. Chrouta, J. Telmoudi, A. J. Zaafouri, A. |
| author_facet | Lachheb, A. Chrouta, J. Telmoudi, A. J. Zaafouri, A. |
| author_institution_txt_mv | [
{
"author": "A. Lachheb",
"institution": "University of Carthage"
},
{
"author": "J. Chrouta",
"institution": "University of Tunis"
},
{
"author": "A. J. Telmoudi",
"institution": "University of Tunis"
},
{
"author": "A. Zaafouri",
"institution": "University of Tunis"
}
] |
| author_sort | Lachheb, A. |
| baseUrl_str | http://eie.khpi.edu.ua/oai |
| collection | OJS |
| datestamp_date | 2026-07-01T21:42:56Z |
| description | Introduction. The increasing penetration of electric vehicles (EVs) and renewable energy has intensified concerns about grid stability and energy sustainability. Integrating photovoltaic (PV) systems with vehicle-to-grid (V2G) technology provides a promising solution but requires efficient energy management and robust control strategies. Problem. Conventional maximum power point tracking (MPPT) methods such as perturb & observe (P&O) suffer from oscillations and poor dynamic response under rapidly changing conditions. Likewise, existing V2G strategies lack adaptive management for optimal renewable utilization and battery protection. Goal. To design an intelligent hybrid control system that maximizes PV power extraction and optimizes EV charging/discharging while ensuring grid stability and extending battery lifespan. Methodology. A two-level hierarchical control architecture is developed. At the low level, an artificial neural network combined with terminal sliding mode control (ANN-TSMC) performs adaptive MPPT. At the high level, a fuzzy logic controller (FLC) manages charging/discharging cycles based on state of charge, grid demand and parking duration. The proposed framework is validated through MATLAB/Simulink simulations. Results. Compared to conventional P&O, the ANN-TSMC controller improves tracking efficiency by 3.6 %, achieves faster convergence (0.14 s), and reduces steady-state oscillations. The FLC reduces grid reliance by 20 % while maintaining a high charging efficiency of 94 %. Furthermore, optimized charging cycles extend battery lifespan by 18.5 %. Scientific novelty. Unlike previous studies limited to single-level control or computationally intensive optimization, this work combines ANN learning ability with TSMC robustness and integrates FLC-based adaptive energy management. Practical value. The proposed system enables resilient PV-based V2G charging stations, reducing grid dependence, improving renewable penetration, and enhancing battery lifetime. These findings support the development of sustainable and grid-friendly EV infrastructures. References 31, tables 2, figures 7. |
| doi_str_mv | 10.20998/2074-272X.2026.4.04 |
| first_indexed | 2026-07-02T01:00:28Z |
| format | Article |
| fulltext |
Electrical Engineering & Electromechanics, 2026, no. 4 25
© A. Lachheb, J. Chrouta, A.J. Telmoudi, A. Zaafouri
UDC 621.311 https://doi.org/10.20998/2074-272X.2026.4.04
A. Lachheb, J. Chrouta, A.J. Telmoudi, A. Zaafouri
Advanced intelligent control for photovoltaic-vehicle-to-grid integration
Introduction. The increasing penetration of electric vehicles (EVs) and renewable energy has intensified concerns about grid stability and
energy sustainability. Integrating photovoltaic (PV) systems with vehicle-to-grid (V2G) technology provides a promising solution but
requires efficient energy management and robust control strategies. Problem. Conventional maximum power point tracking (MPPT)
methods such as perturb & observe (P&O) suffer from oscillations and poor dynamic response under rapidly changing conditions. Likewise,
existing V2G strategies lack adaptive management for optimal renewable utilization and battery protection. Goal. To design an intelligent
hybrid control system that maximizes PV power extraction and optimizes EV charging/discharging while ensuring grid stability and
extending battery lifespan. Methodology. A two-level hierarchical control architecture is developed. At the low level, an artificial neural
network combined with terminal sliding mode control (ANN-TSMC) performs adaptive MPPT. At the high level, a fuzzy logic controller
(FLC) manages charging/discharging cycles based on state of charge, grid demand and parking duration. The proposed framework is
validated through MATLAB/Simulink simulations. Results. Compared to conventional P&O, the ANN-TSMC controller improves tracking
efficiency by 3.6 %, achieves faster convergence (0.14 s), and reduces steady-state oscillations. The FLC reduces grid reliance by 20 %
while maintaining a high charging efficiency of 94 %. Furthermore, optimized charging cycles extend battery lifespan by 18.5 %. Scientific
novelty. Unlike previous studies limited to single-level control or computationally intensive optimization, this work combines ANN learning
ability with TSMC robustness and integrates FLC-based adaptive energy management. Practical value. The proposed system enables
resilient PV-based V2G charging stations, reducing grid dependence, improving renewable penetration, and enhancing battery lifetime.
These findings support the development of sustainable and grid-friendly EV infrastructures. References 31, tables 2, figures 7.
Key words: artificial neural network, terminal sliding mode control, fuzzy logic, maximum power extraction, renewable energy.
Вступ. Зростання поширення електромобілів (EV) та відновлюваних джерел енергії посилило побоювання щодо стабільності
енергомережі та енергетичної стійкості. Інтеграція фотоелектричних (PV) систем з технологією «автомобіль-мережа» (V2G)
є перспективним рішенням, але потребує ефективного управління енергією та надійних стратегій управління. Проблема.
Традиційні методи відстеження точки максимальної потужності (MPPT), такі як метод збурення та спостереження (P&O),
страждають від коливань та поганої динамічної реакції у мінливих умовах. Аналогічно, існуючі стратегії V2G не мають
адаптивного керування для оптимального використання відновлюваних джерел енергії та захисту батарей. Мета. Розробити
інтелектуальну гібридну систему управління, яка максимізує вилучення енергії з PV систем та оптимізує зарядку/розрядку EV,
забезпечуючи при цьому стабільність енергомережі та продовжуючи термін служби батарей. Методика. Розроблено
дворівневу ієрархічну архітектуру управління. На нижньому рівні штучна нейронна мережа у поєднанні з термінальним ковзним
режимом управління (ANN-TSMC) виконує адаптивне MPPT. На верхньому рівні контролер нечіткої логіки (FLC) управляє
циклами зарядки/розрядки на основі стану заряду, попиту в мережі та тривалості стоянки. Запропонована структура
перевірена за допомогою моделювання у MATLAB/Simulink. Результати. У порівнянні з традиційним P&O, контролер ANN-TSMC
підвищує ефективність відстеження на 3,6 %, забезпечує більш швидку збіжність (0,14 с) і знижує коливання в режимі, що
встановився. FLC знижує залежність від мережі на 20 % за збереження високої ефективності зарядки 94%. Крім того,
оптимізовані цикли заряджання збільшують термін служби батареї на 18,5 %. Наукова новизна. На відміну від попередніх
досліджень, обмежених однорівневим керуванням або обчислювально складною оптимізацією, у цій роботі поєднуються
можливості навчання ANN із стійкістю TSMC та інтегрується адаптивне керування енергією на основі FLC. Практична
значимість. Запропонована система дозволяє створювати відмовостійкі зарядні станції V2G на основі PV систем, знижуючи
залежність від мережі, підвищуючи частку відновлюваних джерел енергії та збільшуючи термін служби батареї. Ці результати
сприяють розвитку стійкої та енергозберігаючої інфраструктури для EV. Бібл. 31, табл. 2, рис. 7.
Ключові слова: штучна нейронна мережа, термінальне керування у ковзному режимі, нечітка логіка, відстеження
точки максимальної потужності, відновлювана енергетика.
Introduction. The global transition toward electric
vehicles (EVs) represents a critical pathway to achieving
carbon neutrality, with EV adoption projected to reach 50 %
by 2030. However, this rapid electrification presents
significant challenges to grid stability, primarily due to
the increased and often concentrated energy demand from
charging infrastructure. The integration of photovoltaic
(PV) systems with vehicle-to-grid (V2G) technology
offers a promising solution, yet requires intelligent control
strategies to manage bidirectional power flow, maximize
renewable energy utilization, and protect battery health.
Problem definition and substantiation of its
relevance. The global transition to EVs is a cornerstone
of strategies to reduce fossil fuel reliance and cut CO2
emissions. However, the widespread adoption of EVs
introduces significant challenges for power grids,
primarily due to the increased and often concentrated
energy demand from charging infrastructure.
Simultaneously, the integration of intermittent renewable
energy sources, like solar PV systems, is crucial for
sustainability but poses challenges to grid stability.
Integrating PV systems with V2G technology presents
a promising solution, enabling EVs to act as distributed
energy storage resources. This synergy can enhance grid
stability, improve renewable energy utilization, and provide
economic benefits to EV owners. However, this potential is
hampered by 2 core technical problems:
1. Inefficient maximum power point tracking (MPPT).
Conventional MPPT methods (perturb & observe (P&O),
incremental inductance) suffer from oscillations around the
maximum power point (MPP) and poor dynamic response
under rapidly changing environmental conditions (cloud
cover, partial shading). This leads to significant energy
losses and suboptimal harvesting of available solar energy.
2. Non-adaptive V2G energy management: existing
V2G strategies often lack intelligent, multi-objective
management. They fail to adaptively optimize
charging/discharging cycles based on real-time factors
such as grid demand, EV battery state of charge (SOC),
parking duration, and battery health. These results in
suboptimal grid support, increased grid reliance, and
accelerated battery degradation.
Therefore, developing an intelligent, robust, and
adaptive control system is highly relevant to realizing the
full benefits of PV-based V2G systems, ensuring both
grid stability and the long-term viability of EV batteries.
26 Electrical Engineering & Electromechanics, 2026, no. 4
Review of recent publications with selection of
unsolved tasks. Global carbon-neutral initiatives are driving
the adoption of EVs as a key solution to reduce transport
emissions, especially when powered by renewable energy [1–
4]. This has spurred significant EV research over the past
decade, highlighting the need to integrate renewable energy
into power grids. Despite high initial costs, maximizing EV
system efficiency is crucial. While EV charging methods
have been extensively studied, V2G technology for grid
power remains an emerging research area. While existing
research focuses on V2G/G2V bidirectional energy transfer,
these systems face challenges in maintaining power quality
during AC grid fluctuations, limiting their peak shaving and
fossil fuel displacement potential. Additionally, most studies
neglect direct renewable energy integration at charging
stations, missing opportunities to enhance both sustainability
and grid resilience. To surmount these challenges, this study
proposes intelligent charging algorithms designed to support
grid stability, EV charging can be managed in response to
real-time conditions, allowing EVs to supply stored energy
back to the grid during peak demand, while charging is
strategically scheduled for off-peak hours when renewable
energy is plentiful. Furthermore, EV charging stations will be
outfitted with solar panels to generate clean energy on-site.
Battery electric vehicles (BEVs) offer user-friendly
operation, low maintenance and zero-emission mobility,
making them an environmentally sustainable transportation
solution. The proposed research [5–8] specifically examines
Li-ion battery-powered BEVs, which dominate the EV
market due to their high energy density, longevity and
efficiency. However, widespread BEV adoption depends
critically on developing robust charging infrastructure. DC
charging infrastructure offers distinct advantages over AC
systems by eliminating onboard AC-DC conversion,
improving efficiency. Unlike AC charging constrained by
off-peak demand management, DC systems enable direct
renewable energy integration via DC buses, assuring
bidirectional power flow without conversion losses and
continuous charging availability regardless of grid load.
This DC architecture enhances sustainability while
overcoming AC charging limitations. It simplifies grid
interconnection and reduces auxiliary power requirements.
Existing research has developed multiple control strategies
for V2G/G2V operation. Adaptive control with
bidirectional converters, as outlined in [9], effectively
implements the constant current constant voltage charging
method ensures safe and efficient battery charging, while
constant current discharge control regulates the power flow
to the grid and integrates PV systems using incremental
conductance MPPT technique to effectively harvest solar
energy. However, a notable disadvantage is its lack of an
intelligent management strategy, meaning it does not
optimize charging based on factors like parking duration or
grid demand. These points to a gap in the method: it does
not consider multi-variable optimization for a more
comprehensive energy management approach.
The super-twisting sliding mode control (SMC) offers
a superior dynamic response and reduced chattering
compared to conventional SMC [10]. However, it is limited
to G2V mode only, lacks integration with renewable
energy sources, and does not address bidirectional V2G
operations. While this research concentrated on G2V
charging, future investigations could explore the integration
of renewable energy sources to diminish grid dependence
and enhance environmental sustainability.
To address the lack of a management algorithm, in
reference [11] authors proposed a central control system to
optimize EV-grid energy exchange. This intelligent
aggregator utilizes real-time data to schedule optimal
charge/discharge cycles for EVs, taking into account both
economic and environmental factors. By leveraging V2G
technology, the system aims to mitigate battery degradation
costs and alleviate peak grid demand. Centralized approach
may have scalability issues but limit renewable energy
integration. While the results are promising, further
integration of renewable energy sources could enhance the
system’s performance. Research work [12] introduces a
decentralized power management strategy aimed at
reducing voltage fluctuations in grid-connected energy
storage batteries, thereby improving both battery
performance and grid stability. The primary focus of this
scheme is to lower battery charging costs by utilizing time-
of-use tariffs, with a secondary aim of reducing the
batteries’ charging power through energy generated from
PV. However, the study does not incorporate MPPT
techniques, missing the opportunity to optimize renewable
energy harvesting, which could significantly enhance
overall system performance.
Previous research findings underscore the necessity
for the V2G system controller to embody robustness and
intelligence. Fuzzy logic control presents numerous
advantages for this application [13]. Firstly, it can adapt
to variable conditions, such as battery temperature and
driving patterns, thereby improving battery efficiency and
lifespan [14–18]. Moreover, it adeptly manages
uncertainty in sensor data, enhancing robustness against
fluctuations and inaccuracies. Additionally, it optimizes
real-time performance by employing fuzzy rules,
simultaneously emphasizing charging speed and battery
durability [19]. Lastly, it simplifies model complexity by
utilizing straightforward linguistic rules, thereby
streamlining control system design and implementation
but most of the studies mentioned above lack optimization
algorithms for renewable energy maximization.
Authors [20] demonstrate the significant potential of
fuzzy logic controllers (FLCs) for effective
implementation in power management systems. The
promising results reported in their study, along with those
from related works, strongly support the effectiveness of
fuzzy logic control in providing reliable and adaptive
performance, particularly in V2G applications. A notable
limitation of this work is the absence of an optimization
algorithm specifically designed to maximize renewable
energy production, indicating a potential area for
enhancement. Addressing this gap through the
implementation of an optimization algorithm would not
only improve system performance but also provide
measurable verification of the approach.
The operational deployment of a FLC in V2G
applications is effectively demonstrated in [21], which
offers valuable implementation insights. This research
utilizes streamlined fuzzy logic architecture, specifically
implementing a zero-order Sugeno model, to enhance
computational efficiency while maintaining control
accuracy, to facilitate effective two-way control of an
EV’s power output. This methodology represents a
promising strategy for managing electricity flow between
the EV and the grid.
The control framework [22] employs multi-level
power conversion technology across both grid and EV
Electrical Engineering & Electromechanics, 2026, no. 4 27
terminals of charging stations. A combined fuzzy logic and
PI (FL-PI) control architecture is implemented to manage
the 3-level grid-side converter in an EV charging station. To
demonstrate its effectiveness, the performance of the FL-PI
controller is compared to that of a traditional PI controller
and a PI-fuzzy controller under identical conditions. The
results indicate that the FL-PI controller provides superior
performance, while the PI-fuzzy controller still yields
satisfactory outcomes in terms of settling times and minimal
peak overshoot. But this study needs more efficient and
intelligent control strategies for integrating EVs into the
power grid, particularly for grid support functions such as
reactive power support and voltage regulation.
To address the shortage of MPPT algorithms
mentioned in previous research, the literature has reviewed
various maximum power extraction techniques aimed at
enhancing electricity production in PV systems, examining
a range of optimization methods and algorithms.
Authors [23, 24] have proposed integrating an
artificial neural network (ANN) with particle swarm
optimization to enhance the precision and reliability of
energy harvesting in PV systems, even under varying
weather conditions. However, its significant drawback is
its computationally intensive nature, which leads to longer
processing times. This limitation, specifically the
processing delays, restricts its suitability for real-time
implementation, particularly in scenarios involving rapid
environmental changes.
In [25] authors propose an improved approach based
on the P&O method to overcome limitations in
convergence speed and steady-state oscillations. The
enhanced MPPT method refines MPPT by using the
average of the previous three duty cycles, ensuring greater
precision. P&O is a widely adopted technique due to its
simple implementation and low computational
requirements. However, it exhibits disadvantages such as
slow convergence speed and steady-state oscillations
around the MPP. Furthermore, its performance tends to be
poor under rapidly changing environmental conditions.
The work [26] investigates the application of SMC for
MPPT in PV systems, highlighting its robustness and fast
dynamic response under varying environmental conditions.
They provide a thorough examination that classifies and
analyzes various SMC techniques for maximizing power
extraction in both grid-connected and off-grid applications.
As noted in [26], SMC based methods are well recognized
for their robustness to uncertainties and suitability for
controlling nonlinear systems. However, these methods
also present disadvantages, notably the chattering
phenomenon and potentially slow response times,
particularly in traditional implementations.
Research works [27, 28] also presents fundamental
MPPT approaches, including incremental conductance,
P&O, first-order SMC, and linear expression-based SMC,
along with their adaptive variants. Furthermore, the authors
evaluate advanced SMC approaches, including super
twisting SMC, terminal sliding mode control (TSMC), and
methodologies that incorporate AI algorithms. The
traditional SMC method for MPPT often suffers from slow
response times. To address this issue, TSMC has been
developed as an advanced control solution, offering quicker
convergence and improved performance for DC-DC
converter control. TSMC, an advanced control technique
suitable for systems with uncertainties, builds upon
traditional SMC while delivering several advantages.
Specifically, TSMC guarantees that the system attains the
desired state within a finite timeframe, thereby enhancing
speed and efficiency. Moreover, compared to conventional
SMC, TSMC exhibits reduced sensitivity to system
uncertainties and external disturbances [29]. Building on
SMC, the authors [30] proposes an adaptive SMC algorithm,
incorporating a specialized adaptive tracking mechanism
designed for low-energy disturbance environments. The
stability of this control scheme is rigorously analyzed using
the Lyapunov stability theorem.
Identified research gaps. Based on the literature
review, the following critical gaps remain unaddressed:
1. Lack of hierarchical hybrid control. Existing
studies focus on single-level control or computationally
intensive optimization, without combining
complementary control strategies for both MPPT and
energy management.
2. Insufficient real-time adaptability. Most systems lack
adaptive management integrating multiple parameters (SOC,
parking time (PT), grid demand (GD)) for optimal renewable
utilization and battery protection.
3. Limited renewable integration. Many V2G studies
neglect direct renewable energy integration at charging
stations, missing opportunities for enhanced sustainability.
4. Battery lifespan optimization. Insufficient
attention to optimized charging/discharging cycles that
could significantly extend battery lifespan while
maintaining system performance.
5. Performance under dynamic conditions. Limited
validation of control strategies under rapidly changing
environmental conditions requires both fast response and
minimal oscillations.
6. Integrated system approach. Absence of
comprehensive frameworks combining advanced MPPT
(ANN-TSMC) with intelligent energy management in a
unified, practical architecture.
The goal of this work is to design an intelligent
hybrid control system that maximizes PV power
extraction and optimizes EV charging/discharging while
ensuring grid stability and extending battery lifespan.
First, it focuses on maximized PV power extraction
through the implementation of adaptive MPPT that
maintains high efficiency under rapidly changing
environmental conditions with minimal power
oscillations. Second, the system is designed for optimized
EV charging/discharging by developing an intelligent
energy management strategy that successfully balances
renewable utilization, instantaneous grid demand, and the
crucial requirement of battery protection. Third, by
ensuring reliable bidirectional power flow between EVs
and the utility grid, the system contributes to enhanced
grid stability while reducing dependence on conventional,
non-renewable generation sources. Finally, through the
implementation of optimized charge and discharge cycles,
the control strategies are designed to extend battery
lifespan by actively minimizing degradation.
To successfully achieve this main objective, the
following specific goals are defined. For the low-level
control (MPPT), the first goal is to develop an ANN-TSMC
hybrid controller that leverages the learning ability of
ANNs with the robustness of TSMC. Concurrently, for
high-level control (energy management), the goal is to
design the FLC that integrates real-time data on the
battery’s SOC, parking duration, and grid demand. This
FLC is crucial for optimizing charging/discharging
28 Electrical Engineering & Electromechanics, 2026, no. 4
decisions to maximize renewable energy usage, thereby
reducing grid reliance while maintaining high charging
efficiency and implementing battery protection
mechanisms to extend lifespan. Finally, these elements
must be unified under system integration, which requires
developing a two-level hierarchical control architecture to
ensure seamless coordination between the MPPT and
energy management layers, culminating in the validation of
the complete system through MATLAB/Simulink
simulations to demonstrate practical applicability for real-
world PV-based V2G charging stations.
Design of V2G and G2V installation. Integrating
renewable energy with EV charging stations offers
sustainable benefits like lower carbon emissions, energy
independence, and grid stability. Using solar power cuts
transport emissions and can save costs over time. Solar-
powered EV charging stations cut transport emissions,
boost energy independence, and stabilize the grid. Locally
generated electricity also lowers costs. This renewable EV
synergy supports sustainability goals, regulatory
compliance, and corporate responsibility. Energy storage
is optimized through V2G and G2V technologies,
allowing EVs to charge or discharge power back to the
grid. This dynamic energy exchange helps balance supply
and demand in real time.
This paper presents a hybrid system that integrates PV
arrays with a DC power bus, enabling direct connection to
DC fast charging stations for EVs. A 3-phase voltage
source inverter links the DC bus to the AC grid, facilitating
controlled power transfer for EV charging (Fig. 1). The
control architecture supports both V2G and G2V
operations within a unified framework.
The power conversion system uses a dual-stage
design: a DC-DC flyback converter for voltage regulation
and a voltage source converter with space vector
modulation for efficient DC-AC conversion. This setup
allows dynamic power sharing between the grid and
connected batteries. The proposed system combines V2G
technology with a PV-powered charging station, forming
a hybrid energy management system. The bidirectional
V2G capability enables G2V charging during low-
demand periods, V2G power injection during peak
demand to enhance grid stability and reduce dependence
on conventional generation.
Fig. 1. Hierarchical control system for the V2G/G2V system
Additionally, the EV battery acts as a storage buffer
for excess solar energy, which can either charge the
vehicle or supply power back to the grid as needed. The
PV system utilizes a MPPT algorithm to maximize energy
harvest under varying conditions. A fuzzy logic-based
battery management system dynamically adjusts the SOC,
optimizing charging based on grid demand, parking
duration, and battery health.
For enhanced performance, the PV system
incorporates a hybrid ANN with TSMC, continuously
adapting to environmental changes to maximize power
output. The EV battery pack interfaces with the DC bus via
a bidirectional buck-boost converter, which regulates the
voltage between the battery and the bus to ensure efficient
charging, independent of the battery’s charge level.
The system is designed with 2 primary levels:
1) High-level supervisory control: the FLC. The
FLC acts as the intelligent decision-making core of the
entire charging system. Its primary role is strategic energy
management, not direct power conversion control.
2) Low-level execution control: the ANN-TSMC
hybrid controller. This hybrid controller is responsible for
the fast, real-time control of the power converter. It is
activated and guided by the high-level FLC. Its main
purpose is to execute the command from the FLC with
high efficiency and robustness, specifically for the MPPT
of the PV array.
Design of an adaptive robust MPPT control
strategy. This paper proposes a hybrid ANN-TSMC
control strategy to achieve adaptive MPPT in PV systems
under rapidly varying environmental conditions. The
ANN algorithm generates an optimal voltage, while the
TSMC component ensures fast convergence and robust
tracking despite irradiance fluctuations and partial
shading effects.
Reference voltage generation by ANN algorithm.
The ANN algorithm generates an optimal reference
voltage by learning the nonlinear characteristics of the PV
array. The network is trained on systematically collected
data, capturing variations in temperature and irradiance,
which are correlated with the corresponding MPP
voltages. This enables the ANN to accurately model and
predict optimal operating conditions for power
maximization [31].
The neural network employs a simple three-layer
architecture: an input layer receiving solar temperature
and irradiance data, a hidden layer for internal processing
consists of 10 neurons with a sigmoid activation function,
and an output layer delivering the final result consist of 1
neuron (VMPP) with a linear activation function. Neurons
within each layer are interconnected via weighted
connections and employ activation functions. The ANN
used in this study was trained on a dataset of 500 data
points, each consisting of irradiance (G) and temperature
(T) as inputs, and the corresponding MPP voltage (VMPP)
as the output. The data was collected from a real-world
PV system under a wide range of environmental
conditions, with irradiance ranging from 200-1000 W/m2
in 100 W/m2 steps and temperature ranging from 15 °C to
55 °C in 5 °C increments. The dataset was preprocessed
to remove outliers and normalized to ensure consistency.
The ANN was trained via the Levenberg-Marquardt back
propagation algorithm, with the dataset partitioned into 65 %
training, 15 % validation, and 20 % testing subsets to
ensure robust performance evaluation. By training the
ANN on this diverse dataset, the model learns to
accurately predict the MPP voltage (VMPP) under a wide
range of environmental conditions, ensuring robust
Electrical Engineering & Electromechanics, 2026, no. 4 29
performance in real-world applications. The ANN
achieved an R2 value of 0.98 on the test set,
demonstrating its ability to accurately predict VMPP under
varying environmental conditions.
To further validate the ANN’s performance,
additional simulations were conducted under varying
environmental conditions. The finding demonstrated that
the ANN accurately predicts the MPP voltage (VMPP)
across the entire range of irradiance and temperature
values, ensuring reliable PV system operation. Given that
solar panel characteristics are influenced by weather
conditions, the neural network utilizes irradiation and
solar module temperature as inputs. Figure 2 illustrates
the feed forward network architecture.
1e
2e
*
pvV
*
Li Li
Li
pvV
Fig. 2. Schematic of the flyback converter implementing
the proposed MPPT control
The proposed ANN accurately predicts the voltage
at the MPP of the PV module. This estimated value is
used as the reference voltage (Vpv
*) for the TSMC
controller. The control system minimizes the error
e1 = Vpv – Vpv
*, where Vpv is the actual PV voltage, to drive
the system to operate at VMPP. By ensuring that Vpv
converges to Vpv
*, the system achieves MPPT.
Terminal sliding mode control (TSMC). SMC forces
the system’s state trajectory to converge onto a designed
sliding manifold in finite time via discontinuous control.
TSMC uses PV voltage, reference voltage, and current
changes to generate the duty ratio output. The control design
process commences with the definition of tracking error
variables as follows. The control design process commences
with the definition of tracking error variables as:
e1 = Vpv – Vpv
*, (1)
where e1 is the voltage error signal; Vpv is the measured
PV voltage; Vpv
* is the reference PV voltage.
The current tracking error is described as:
e2 = iL – iL
*, (2)
where iL is the inductor current; iL
* is the inductor current
reference.
The inductor current reference iL
* is derived from the
PV system dynamics, the relationship between the PV
current ipv, the inductor current iL and the PV voltage Vpv:
t
V
Cii pv
pvpvL d
d
. (3)
The current reference is adopted as:
iL
* = ipv – CpvVpv
*.
By incorporating the expression of the current
reference into (2), the current tracking error e2 becomes:
e2 = iL – ipv + CpvVpv
*. (4)
The error dynamics are derived through the
following analytical procedure:
pv
pvLpv
pv C
e
Vii
C
e 2*
1
1
; (5)
*
2 1
11
Ldcpv ituV
L
V
L
e , (6)
where 1e is the time derivative of the voltage error signal;
Cₚᵥ is the PV capacitance; 2e is the time derivative of the
current error signal; L is the inductance; Vdc is the DC-
link voltage; u is the control input (duty cycle); δ(t) is the
system uncertainty or disturbance.
The proposed control system is designed to
simultaneously achieve 2 primary objectives:
1) ensuring zero convergence of tracking errors (1)
and (2) within a finite time for precise tracking;
2) the PWM control signals are synthesized to
enforce the reaching condition SṠ < 0.
The TSMC controller guarantees finite-time
convergence of the system trajectory S to the sliding
manifold, even under matched uncertainties. This is
achieved through a co-designed nonlinear sliding
manifold and a discontinuous control law with fractional-
power terms, ensuring Lyapunov stability and
deterministic bounded-time convergence.
For robust MPPT, we use SMC with surface (7),
ensuring the system reaches and maintains the desired
trajectory by enforcing S =0. The terminal sliding surface
is defined:
12
1
eeS x
, (7)
where S is the terminal sliding manifold in the TSMC
framework, used to ensure finite-time convergence and
robust control of the PV system; >0, with x = p / q and
1<x<2, the values for p and q should be positive odd
numbers, as this helps meet specific mathematical
requirements that ensuring system stability and reliability
under the specified constraints 0<q<p [31].
When the sliding condition is achieved S(t)=0, the
current error becomes as:
1
12
1
ee . (8)
The equation of dynamic error describes in (5)
becomes:
x
pv
x
e
C
e 1
1
1
1
. (9)
Once the system reaches and remains on the sliding
surface (S(t) = 0), the errors (e1, e2) are driven to 0. To
ensure reachability of this surface, a framework is
employed that yields the intended outcomes:
Theorem. The equations (5), (6) describe the
dynamic behavior of a PV system. Applying robust
TSMC implemented with a specific control law u(t)
developed to force the system reaches the sliding surface
S=0 within a finite time interval and guarantees tracking
of the maximum power:
Si
xC
e
VV
L
V
tu
L
pv
x
dcpvdc sign
1
1
*
2
2
. (10)
30 Electrical Engineering & Electromechanics, 2026, no. 4
The parameters α>0 and σ>0 ensure the reachability
of the sliding surface and robust control under
uncertainties. The finite-time convergence of the system
to the sliding surface is rigorously verified using the
Lyapunov stability criterion:
V(S) = 0.5S2. (11)
Applying control rule (10), the time derivative of
V(S) can be expressed:
SeeSSSSV x
22
1
. (12)
Control strategy for EV batteries. Existing research
explores various battery charging methodologies, including
constant current (CC), constant voltage (CV), and hybrid
CC-CV techniques. Among these, the CC-CV method is
widely recommended by manufacturers for Li-ion batteries,
as it optimally balances charging efficiency and battery
lifespan. A critical aspect of this approach is the precise
current regulation of buck converters, which ensures
controlled energy transfer during both charging and
discharging phases. The proposed CC-CV control strategy,
illustrated in Fig. 3, integrates adaptive current and voltage
regulation to enhance battery performance while mitigating
degradation. This design not only adheres to industry
standards but also improves upon conventional methods by
dynamically adjusting to real-time load and SOC conditions.
Fig. 3. Configuration of the V2G/G2V
Fuzzy logic-based EV charging management can
optimize grid integration of recovered and renewable
energy while extending battery life. The proposed
controller uses three inputs: SOC battery, parking
duration, and current grid demand.
The FLC output determines whether to charge,
discharge, or maintain the EV battery’s current state. It
employs triangular membership functions for each input
variable, chosen for their balance of simplicity and
accuracy, and uses a rule set to determine appropriate
actions based on inputs.
FLC design. The FLC aims to optimize the charging
and discharging of the EV battery based on the 3 inputs.
The design process comprises these steps:
1. Fuzzification. The input values (SOC, PT, GD)
are fuzzified into fuzzy sets with the aid of triangular
membership functions. These functions define the degree
to which each input belongs to a specific category (low,
medium, high).
2. Rule base. The fuzzy logic rules are designed to
optimize the charging of the EV battery based on the
inputs (SOC, PT and GD). The FLC employs a set of 27
rules to make the appropriate action (charge, discharge or
maintain).
3. Inference engine. The inference mechanism
assesses the rules and determines the degree to which
each rule applies based on the input values. A fuzzy
output is generated by combining the rules using the
Mamdani inference method.
4. Defuzzification. The centroid method is used to
defuzzify the fuzzy output into a crisp value, which
calculates the center of gravity of the output membership
function. The crisp output determines the action to be
taken (charge, discharge or maintain).
The membership functions designed for the
proposed system are illustrated in Fig. 4.
GD
PT, s
SOC, %
Fig. 4. Membership functions of the fuzzy sets for each input
The FLC is pivotal in determining the optimal
charge/discharge actions for EV batteries. By dynamically
adapting to real-time conditions, the FLC enhances system
performance through 3 key objectives: 1) maintaining grid
stability; 2) prolonging battery lifespan; 3) optimizing user
convenience.
Results and discussions. MPPT performance
evaluation. This section evaluates the proposed ANN-
TSMC MPPT controllers under various operating
scenarios, comparing them with traditional P&O methods
(Fig. 5).
t, s
P, W
Fig. 5. PV power generation considering constant temperature
and variable irradiation
Electrical Engineering & Electromechanics, 2026, no. 4 31
When irradiance abruptly increased from 850 W/m2
to 1000 W/m2 at t = 1 s, PV power output rose from 2 kW
to 2.5 kW. Results show P&O algorithms generate less
power with larger oscillations than the proposed method,
which demonstrates superior dynamic response,
particularly during transient states. The ANN-TSMC
approach effectively addresses the significant MPPT
divergence exhibited by P&O methods during rapid
condition changes.
Table 1 compares performance metrics for different
MPPT approaches, evaluating response time, oscillation,
efficiency, and energy losses. The proposed ANN-TSMC
method clearly outperforms traditional P&O, particularly
in efficiency and stability.
Under conditions of rapid irradiance changes (from
800 to 1000 W/m2), the control methods demonstrated
significantly different levels of effectiveness. The
proposed ANN-TSMC method proved to be the most
robust, maintaining a high efficiency of 97.8 % with only
minimal power loss during the transitions. In comparison,
the ANN-PSO method also achieved high accuracy at
96.8 %, but its practical application was limited by a 57 %
slower response time due to computational overhead. The
incremental conductance method delivered moderate
performance, reaching 95.6 % efficiency, though it was
prone to occasional tracking errors. The P&O method
struggled the most under these dynamic conditions, with
its efficiency dropping sharply to 91.5 % as it experienced
significant tracking delays. This enhanced technique
demonstrates superior performance and robustness under
varying weather conditions, making it a promising control
solution for solar systems during MPPT operation.
Table 1
Comprehensive MPPT algorithm performance comparison
Method
Tracking
efficiency, %
Settling time, s Overshoot, %
Power
oscillation, W
Computational
complexity
Real-time
suitability
Proposed ANN-TSMC 97.8 0.14 3 ±15 Medium Excellent
P&O 91.5 0.18 12 ±45 Low Good
Incremental conductance 95.6 0.16 8 ±35 Medium Good
ANN-PSO [13, 14] 96.8 0.22 5 ±25 Very High Poor
Fuzzy logic MPPT 95.1 0.19 10 ±40 Medium Good
Traditional SMC 96.2 0.17 7 ±30 Medium Fair
Energy management performance. To validate the
functionality of the proposed FLC management system,
different simulations were conducted under various
scenarios. The different scenarios are simulated for the
same weather conditions and the same load.
1st scenario. Storage system (EV1) begins fully
discharged with no EV2 integration. Initially, without
sunlight, the grid supplied all load demands as the PV
system generated no power. As PV power gradually
increased to peak at approximately 7500 W at 10:00 h,
EV2’s battery was able to charge. Figure 6 illustrates the
power generation in this 1st scenario.
t, h
P, W
Fig. 6. PV system power for the 1st scenario
2nd scenario. Vehicle battery is charged, and
storage battery discharged. In this case, the following
conditions were considered: Assuming the vehicle is
parked for over 3 h, and the EV battery is fully charged,
and the storage batteries are completely discharged. The
EV battery began to discharge to provide energy to the
load aiding in stabilizing electricity demand during peak
periods. Figure 7 offers a comprehensive visualization of
the power dynamics in the 2nd scenario, showing how the
system adjusts to demand and irradiance fluctuations, and
effectively distributes power among the PV system,
battery storage, EVs and grid.
t, h
P, W
Fig. 7. Power of grid connected PV system for the 2nd scenario
Analysis of these power curves demonstrates the
system’s autonomous operation and ability to satisfy
energy demand. Effective battery management provides
system resilience and minimizes external power
dependence. Both the EV and storage system batteries are
crucial for maintaining equilibrium by storing excess
energy during surplus periods and releasing it during peak
demand, reducing reliance on the electricity grid. FLCs
provide key advantages for EV charging: adaptability to
dynamic conditions like fluctuating grid demand and
battery state-of-charge, robustness in handling sensor data
uncertainties, efficiency through real-time optimization
balancing charging speed and battery durability and
simplified implementation using linguistic rules that
reduce model complexity. Table 2 gives a comparison
between FLC and existing control methods for EV battery
charging applications.
32 Electrical Engineering & Electromechanics, 2026, no. 4
Table 2
Comprehensive V2G/G2V controller comparison
Control
method
Charging
efficiency,
%
Response
time, s
Battery
life
extension,
%
Grid
stability
index
Implementation
complexity
Fuzzy
logic [27]
94.5 0.12 18.5 0.95 Medium
FL-PI
[29]
89.2 0.28 8.2 0.78 Low
Adaptive
control
[9]
91.8 0.22 12.1 0.82 High
Central
control
[11]
90.6 0.35 10.8 0.85 Very High
FLCs can reduce charging time by dynamically
optimizing charging profiles, maximizing current while
minimizing battery stress. By adapting to real-time
conditions like SOC and temperature, they enable faster
charging without compromising safety. Their flexibility
allows dynamic charging curve adjustments in variable
scenarios. In contrast, PI regulators deliver consistent but
potentially slower charging times within design limits,
lacking adaptability to fluctuations in grid demand or
battery parameters.
The proposed FLC demonstrates superior decision-
making capabilities by considering three inputs, SOC,
parking duration, and grid demand and utilizing 27
optimized rules. This multi-objective approach surpasses
conventional methods, which are typically limited to
single or dual objectives, and also improves upon other
multi-objective strategies like the one in reference [26],
which lacks real-time adaptability. The practical benefits of
this advanced logic are evident in operational scenarios.
For instance, in peak demand management, the proposed
system achieves a 23 % reduction in peak load through
intelligent V2G scheduling, significantly outperforming the
8–15 % reduction from conventional systems and the 18 %
from advanced centralized controls [26], which also suffer
from communication overhead. Furthermore, when
assessing renewable energy integration, the proposed
system attains 89 % utilization efficiency, a substantial
improvement over the 65–75 % efficiency of basic
integration methods and the complete grid dependence of
systems without any integration.
The V2G/G2V operations can greatly benefit from
integrating smart charging algorithms with renewable
energy sources. Smart algorithms optimize EV charging
based on real-time conditions, EVs discharge during peak
demand to reduce strain on power plants and charge
during off-peak hours when renewables are abundant.
Installing renewable energy sources at charging stations
provides clean, on-site power for EVs that can also feed
back to the grid during highest demand periods.
Advanced grid management systems enable coordination
between EVs, charging infrastructure, renewables, and the
main grid, optimizing energy flow and maintaining
stability despite renewable energy fluctuations. Benefits
include reduced peak demand, improved grid stability,
and enhanced sustainability through a cleaner energy
ecosystem. This integrated approach addresses V2G/G2V
limitations on AC grids while creating more sustainable,
resilient, and efficient EV charging infrastructure.
Conclusions. This study presented a smart charging
framework integrating renewable energy-powered
V2G/G2V systems with advanced control algorithms to
achieve sustainable and efficient EV charging
infrastructure. A hybrid ANN-TSMC MPPT control,
ensures maximum power extraction from PV systems
under dynamic environmental conditions, simulation
results demonstrate 3.6 % higher energy capture compared
to conventional P&O methods, significantly improving
renewable energy utilization in V2G operations.
The intelligent charging control combines a FLC with
constant current constant voltage to optimize EV battery
charging. The FLC tracks SOC battery, parking time, and
grid demand to adapt charging strategies in real-time.
During peak demand, the system prioritizes battery
discharge to support the grid and reduce reliance on
conventional power. The FLC prevents overcharging and
excessive discharging to extend battery life while ensuring
adequate charge by scheduled departure times. This
intelligent control led to a 20 % reduction in reliance on
conventional power, achieved a charging efficiency of
94 %, and operated with a fast response time of 0.12 s.
Crucially, by preventing overcharging and excessive
discharging, the system contributes to an estimated 18.5 %
extension in battery lifespan.
To further advance this research, we propose two main
directions: first, hardware-in-the-loop validation to test the
system’s robustness under extreme weather conditions, and
second, multi-objective optimization for large-scale EV
fleets to address grid congestion and ensure fairness in
energy allocation. In summary, this research demonstrates
the clear advantages of integrating smart algorithms with
renewables in V2G/G2V systems, contributing to a resilient
and sustainable EV charging infrastructure.
Conflict of interest. The authors declare that they
have no conflicts of interest.
REFERENCES
1. Kumar P., Channi H.K., Rani S., Kataria A., Rattan P. Enhanced
grid integration through machine-learning optimized bidirectional EV
chargers. Sustainable Energy, Grids and Networks, 2025, vol. 44, art.
no. 102007. doi: https://doi.org/10.1016/j.segan.2025.102007.
2. Goel S., Sharma R., Rathore A.K. A review on barrier and
challenges of electric vehicle in India and vehicle to grid optimisation.
Transportation Engineering, 2021, vol. 4, art. no. 100057. doi:
https://doi.org/10.1016/j.treng.2021.100057.
3. Calise F., Fabozzi S., Vanoli L., Vicidomini M. A sustainable
mobility strategy based on electric vehicles and photovoltaic panels
for shopping centers. Sustainable Cities and Society, 2021, vol. 70,
art. no. 102891. doi: https://doi.org/10.1016/j.scs.2021.102891.
4. Grechko O., Kulyk O. Current State and Future Prospects of
Using SF6 Gas as an Insulation in the Electric Power Industry. 2024
IEEE 5th KhPI Week on Advanced Technology (KhPIWeek), 2024,
pp. 1-6. doi: https://doi.org/10.1109/KhPIWeek61434.2024.10877987.
5. Liao J.-T., Huang H.-W., Yang H.-T., Li D. Decentralized
V2G/G2V Scheduling of EV Charging Stations by Considering the
Conversion Efficiency of Bidirectional Chargers. Energies, 2021, vol.
14, no. 4, art. no. 962. doi: https://doi.org/10.3390/en14040962.
6. Lu L., Han X., Li J., Hua J., Ouyang M. A review on the key
issues for lithium-ion battery management in electric vehicles. Journal
of Power Sources, 2013, vol. 226, pp. 272-288. doi:
https://doi.org/10.1016/j.jpowsour.2012.10.060.
7. Chen S., Dai F., Cai M. Opportunities and Challenges of High-
Energy Lithium Metal Batteries for Electric Vehicle Applications.
ACS Energy Letters, 2020, vol. 5, no. 10, pp. 3140-3151. doi:
https://doi.org/10.1021/acsenergylett.0c01545.
Electrical Engineering & Electromechanics, 2026, no. 4 33
8. Rangarajan S.S., Sunddararaj S.P., Sudhakar A., Shiva C.K.,
Subramaniam U., Collins E.R., Senjyu T. Lithium-Ion Batteries – The
Crux of Electric Vehicles with Opportunities and Challenges. Clean
Technologies, 2022, vol. 4, no. 4, pp. 908-930. doi:
https://doi.org/10.3390/cleantechnol4040056.
9. Singirikonda S., Obulesu Y.P., Kannan R., Reddy K.J., Kiran
Kumar G., Alhakami W., Baz A., Alhakami H. Adaptive control-
based Isolated bi-directional converter for G2V& V2G charging with
integration of the renewable energy source. Energy Reports, 2022,
vol. 8, pp. 11416-11428. doi:
https://doi.org/10.1016/j.egyr.2022.08.223.
10. Ahmed I., Ahmad I., Ahmed S., Adil H.M.M. Robust nonlinear
control of battery electric vehicle charger in grid to vehicle
applications. Journal of Energy Storage, 2021, vol. 42, art. no.
103039. doi: https://doi.org/10.1016/j.est.2021.103039.
11. Makeen P., Ghali H.A., Memon S., Duan F. Insightful Electric
Vehicle Utility Grid Aggregator Methodology Based on the G2V and
V2G Technologies in Egypt. Sustainability, 2023, vol. 15, no. 2, art.
no. 1283. doi: https://doi.org/10.3390/su15021283.
12. Al Essa M.J.M. Power management of grid-integrated energy
storage batteries with intermittent renewables. Journal of Energy
Storage, 2020, vol. 31, art. no. 101762. doi:
https://doi.org/10.1016/j.est.2020.101762.
13. De Luca F., Calderaro V., Galdi V. A Fuzzy Logic-Based Control
Algorithm for the Recharge/V2G of a Nine-Phase Integrated On-
Board Battery Charger. Electronics, 2020, vol. 9, no. 6, art. no. 946.
doi: https://doi.org/10.3390/electronics9060946.
14. Vivas F.J., Segura F., Andújar J.M., Palacio A., Saenz J.L., Isorna
F., López E. Multi-Objective Fuzzy Logic-Based Energy
Management System for Microgrids with Battery and Hydrogen
Energy Storage System. Electronics, 2020, vol. 9, no. 7, art. no. 1074.
doi: https://doi.org/10.3390/electronics9071074.
15. Chaib H., Hassaine S., Mihoub Y., Moreau S. Intelligent power
control strategy based on self-tuning fuzzy MPPT for grid-connected
hybrid system. Electrical Engineering & Electromechanics, 2025, no.
3, pp. 23-30. doi: https://doi.org/10.20998/2074-272X.2025.3.04.
16. Lachheb A., Chrouta J., Akoubi N., Ben Salem J. Enhancing
efficiency and sustainability: a combined approach of ANN-based
MPPT and fuzzy logic EMS for grid-connected PV systems. Euro-
Mediterranean Journal for Environmental Integration, 2025, vol. 10,
no. 4, pp. 3067-3079. doi: https://doi.org/10.1007/s41207-024-00727-5.
17. Vivas F.J., Segura F., Andújar J.M., Palacio A., Saenz J.L., Isorna
F., López E. Multi-Objective Fuzzy Logic-Based Energy
Management System for Microgrids with Battery and Hydrogen
Energy Storage System. Electronics, 2020, vol. 9, no. 7, art. no. 1074.
doi: https://doi.org/10.3390/electronics9071074.
18. Hadjidj N., Benbrahim M., Ounnas D., Mouss L.H. Global
maximum power point tracking method for photovoltaic systems
using Takagi-Sugeno fuzzy models and ANFIS approach. Electrical
Engineering & Electromechanics, 2025, no. 2, pp. 31-38. doi:
https://doi.org/10.20998/2074-272X.2025.2.05.
19. Djafer L., Taleb R., Mehedi F., Aissa Bokhtache A., Bessaad T.,
Chabni F., Saidi H. Electric drive vehicle based on sliding mode
control technique using a 21-level asymmetrical inverter under
different operating conditions. Electrical Engineering &
Electromechanics, 2025, no. 3, pp. 31-36. doi:
https://doi.org/10.20998/2074-272X.2025.3.05.
20. Nag S., Lee K.Y. Optimized Fuzzy Logic Controller for
Responsive Charging of Electric Vehicles. IFAC-PapersOnLine,
2019, vol. 52, no. 4, pp. 147-152. doi:
https://doi.org/10.1016/j.ifacol.2019.08.170.
21. Rahmani-Andebili M. Applications of Fuzzy Logic in Planning
and Operation of Smart Grids. Springer Publ., 2021. 229 p. doi:
https://doi.org/10.1007/978-3-030-64627-1.
22. Sayed K., Gabbar H. Electric Vehicle to Power Grid Integration
Using Three-Phase Three-Level AC/DC Converter and PI-Fuzzy
Controller. Energies, 2016, vol. 9, no. 7, art. no. 532. doi:
https://doi.org/10.3390/en9070532.
23. Al-Majidi S.D., Abbod M.F., Al-Raweshidy H.S. A particle
swarm optimisation-trained feedforward neural network for predicting
the maximum power point of a photovoltaic array. Engineering
Applications of Artificial Intelligence, 2020, vol. 92, art. no. 103688.
doi: https://doi.org/10.1016/j.engappai.2020.103688.
24. Nouri A., Lachheb A., El Amraoui L. Optimizing efficiency of
Vehicle-to-Grid system with intelligent management and ANN-PSO
algorithm for battery electric vehicles. Electric Power Systems
Research, 2024, vol. 226, art. no. 109936. doi:
https://doi.org/10.1016/j.epsr.2023.109936.
25. Saberi A., Niroomand M., Dehkordi B.M. An Improved P&O
Based MPPT for PV Systems with Reduced Steady-State Oscillation.
International Journal of Energy Research, 2023, vol. 2023, pp. 1-15.
doi: https://doi.org/10.1155/2023/4694583.
26. Ahmad F.F., Ghenai C., Hamid A.K., Bettayeb M. Application of
sliding mode control for maximum power point tracking of solar
photovoltaic systems: A comprehensive review. Annual Reviews in
Control, 2020, vol. 49, pp. 173-196. doi:
https://doi.org/10.1016/j.arcontrol.2020.04.011.
27. Makhlouf M., Laouar O. New adaptive modified perturb and
observe algorithm for maximum power point tracking in photovoltaic
systems with interleaved boost converter. Electrical Engineering &
Electromechanics, 2025, no. 6, pp. 57-63. doi:
https://doi.org/10.20998/2074-272X.2025.6.08.
28. Toualbia A., Djilali A.B., Taleb R., Mansour N., Souaihia M.,
Saidi H. Application perturb and observe maximum power point
tracking with interconnection and damping assignment passive-based
control for photovoltaic system using boost converter. Electrical
Engineering & Electromechanics, 2025, no. 6, pp. 98-105. doi:
https://doi.org/10.20998/2074-272X.2025.6.13.
29. Abianeh A.J., Ferdowsi F. Sliding Mode Control Enabled Hybrid
Energy Storage System for Islanded DC Microgrids with Pulsing
Loads. Sustainable Cities and Society, 2021, vol. 73, art. no. 103117.
doi: https://doi.org/10.1016/j.scs.2021.103117.
30. Koofigar H.R. Adaptive robust maximum power point tracking
control for perturbed photovoltaic systems with output voltage
estimation. ISA Transactions, 2016, vol. 60, pp. 285-293. doi:
https://doi.org/10.1016/j.isatra.2015.11.003.
31. Khan Z.A., Khan L., Ahmad S., Mumtaz S., Jafar M., Khan Q.
RBF neural network based backstepping terminal sliding mode MPPT
control technique for PV system. PLOS ONE, 2021, vol. 16, no. 4, art.
no. e0249705. doi: https://doi.org/10.1371/journal.pone.0249705.
Received 05.11.2025
Accepted 28.01.2026
Published 02.07.2026
A. Lachheb 1, Doctor of Electrical Engineering,
J. Chrouta 2, Assistant Professor,
A.J. Telmoudi 2, Professor,
A. Zaafouri 2, Professor,
1 Research Laboratory Smart Electricity & ICT, SE&ICT Lab.,
LR18ES44, National Engineering School of Carthage,
University of Carthage, Tunisia,
e-mail: aymen.lachheb@enicarthage.rnu.tn (Corresponding Author)
2 Laboratory of Industrial Systems Engineering and Renewable
Energy (LISIER), Higher National Engineering School of Tunis
(ENSIT), University of Tunis, Tunisia.
How to cite this article:
Lachheb A., Chrouta J., Telmoudi A.J., Zaafouri A. Advanced intelligent control for photovoltaic-vehicle-to-grid integration. Electrical
Engineering & Electromechanics, 2026, no. 4, pp. 25-33. doi: https://doi.org/10.20998/2074-272X.2026.4.04
|
| id | eiekhpieduua-article-343893 |
| institution | Electrical Engineering & Electromechanics |
| keywords_txt_mv | keywords |
| language | English |
| last_indexed | 2026-07-02T01:00:28Z |
| publishDate | 2026 |
| publisher | National Technical University "Kharkiv Polytechnic Institute" and Аnatolii Pidhornyi Institute of Power Machines and Systems of NAS of Ukraine |
| record_format | ojs |
| resource_txt_mv | eiekhpieduua/37/36b26a4cc35fc2006b130dab0190ff37.pdf |
| spelling | eiekhpieduua-article-3438932026-07-01T21:42:56Z Advanced intelligent control for photovoltaic-vehicle-to-grid integration Advanced intelligent control for photovoltaic-vehicle-to-grid integration Lachheb, A. Chrouta, J. Telmoudi, A. J. Zaafouri, A. artificial neural network terminal sliding mode control fuzzy logic maximum power extraction renewable energy штучна нейронна мережа термінальне керування у ковзному режимі нечітка логіка відстеження точки максимальної потужності відновлювана енергетика Introduction. The increasing penetration of electric vehicles (EVs) and renewable energy has intensified concerns about grid stability and energy sustainability. Integrating photovoltaic (PV) systems with vehicle-to-grid (V2G) technology provides a promising solution but requires efficient energy management and robust control strategies. Problem. Conventional maximum power point tracking (MPPT) methods such as perturb &amp; observe (P&amp;O) suffer from oscillations and poor dynamic response under rapidly changing conditions. Likewise, existing V2G strategies lack adaptive management for optimal renewable utilization and battery protection. Goal. To design an intelligent hybrid control system that maximizes PV power extraction and optimizes EV charging/discharging while ensuring grid stability and extending battery lifespan. Methodology. A two-level hierarchical control architecture is developed. At the low level, an artificial neural network combined with terminal sliding mode control (ANN-TSMC) performs adaptive MPPT. At the high level, a fuzzy logic controller (FLC) manages charging/discharging cycles based on state of charge, grid demand and parking duration. The proposed framework is validated through MATLAB/Simulink simulations. Results. Compared to conventional P&amp;O, the ANN-TSMC controller improves tracking efficiency by 3.6 %, achieves faster convergence (0.14 s), and reduces steady-state oscillations. The FLC reduces grid reliance by 20 % while maintaining a high charging efficiency of 94 %. Furthermore, optimized charging cycles extend battery lifespan by 18.5 %. Scientific novelty. Unlike previous studies limited to single-level control or computationally intensive optimization, this work combines ANN learning ability with TSMC robustness and integrates FLC-based adaptive energy management. Practical value. The proposed system enables resilient PV-based V2G charging stations, reducing grid dependence, improving renewable penetration, and enhancing battery lifetime. These findings support the development of sustainable and grid-friendly EV infrastructures. References 31, tables 2, figures 7. Вступ. Зростання поширення електромобілів (EV) та відновлюваних джерел енергії посилило побоювання щодо стабільності енергомережі та енергетичної стійкості. Інтеграція фотоелектричних (PV) систем з технологією «автомобіль-мережа» (V2G) є перспективним рішенням, але потребує ефективного управління енергією та надійних стратегій управління. Проблема. Традиційні методи відстеження точки максимальної потужності (MPPT), такі як метод збурення та спостереження (P&amp;O), страждають від коливань та поганої динамічної реакції у мінливих умовах. Аналогічно, існуючі стратегії V2G не мають адаптивного керування для оптимального використання відновлюваних джерел енергії та захисту батарей. Мета. Розробити інтелектуальну гібридну систему управління, яка максимізує вилучення енергії з PV систем та оптимізує зарядку/розрядку EV, забезпечуючи при цьому стабільність енергомережі та продовжуючи термін служби батарей. Методика. Розроблено дворівневу ієрархічну архітектуру управління. На нижньому рівні штучна нейронна мережа у поєднанні з термінальним ковзним режимом управління (ANN-TSMC) виконує адаптивне MPPT. На верхньому рівні контролер нечіткої логіки (FLC) управляє циклами зарядки/розрядки на основі стану заряду, попиту в мережі та тривалості стоянки. Запропонована структура перевірена за допомогою моделювання у MATLAB/Simulink. Результати. У порівнянні з традиційним P&amp;O, контролер ANN-TSMC підвищує ефективність відстеження на 3,6 %, забезпечує більш швидку збіжність (0,14 с) і знижує коливання в режимі, що встановився. FLC знижує залежність від мережі на 20 % за збереження високої ефективності зарядки 94%. Крім того, оптимізовані цикли заряджання збільшують термін служби батареї на 18,5 %. Наукова новизна. На відміну від попередніх досліджень, обмежених однорівневим керуванням або обчислювально складною оптимізацією, у цій роботі поєднуються можливості навчання ANN із стійкістю TSMC та інтегрується адаптивне керування енергією на основі FLC. Практична значимість. Запропонована система дозволяє створювати відмовостійкі зарядні станції V2G на основі PV систем, знижуючи залежність від мережі, підвищуючи частку відновлюваних джерел енергії та збільшуючи термін служби батареї. Ці результати сприяють розвитку стійкої та енергозберігаючої інфраструктури для EV. Бібл. 31, табл. 2, рис. 7. National Technical University "Kharkiv Polytechnic Institute" and Аnatolii Pidhornyi Institute of Power Machines and Systems of NAS of Ukraine 2026-07-02 Article Article application/pdf https://eie.khpi.edu.ua/article/view/343893 10.20998/2074-272X.2026.4.04 Electrical Engineering & Electromechanics; No. 4 (2026); 25-33 Электротехника и Электромеханика; № 4 (2026); 25-33 Електротехніка і Електромеханіка; № 4 (2026); 25-33 2309-3404 2074-272X en https://eie.khpi.edu.ua/article/view/343893/351594 Copyright (c) 2025 A. Lachheb, J. Chrouta, A. J. Telmoudi, A. Zaafouri http://creativecommons.org/licenses/by-nc/4.0 |
| spellingShingle | artificial neural network terminal sliding mode control fuzzy logic maximum power extraction renewable energy Lachheb, A. Chrouta, J. Telmoudi, A. J. Zaafouri, A. Advanced intelligent control for photovoltaic-vehicle-to-grid integration |
| title | Advanced intelligent control for photovoltaic-vehicle-to-grid integration |
| title_alt | Advanced intelligent control for photovoltaic-vehicle-to-grid integration |
| title_full | Advanced intelligent control for photovoltaic-vehicle-to-grid integration |
| title_fullStr | Advanced intelligent control for photovoltaic-vehicle-to-grid integration |
| title_full_unstemmed | Advanced intelligent control for photovoltaic-vehicle-to-grid integration |
| title_short | Advanced intelligent control for photovoltaic-vehicle-to-grid integration |
| title_sort | advanced intelligent control for photovoltaic-vehicle-to-grid integration |
| topic | artificial neural network terminal sliding mode control fuzzy logic maximum power extraction renewable energy |
| topic_facet | artificial neural network terminal sliding mode control fuzzy logic maximum power extraction renewable energy штучна нейронна мережа термінальне керування у ковзному режимі нечітка логіка відстеження точки максимальної потужності відновлювана енергетика |
| url | https://eie.khpi.edu.ua/article/view/343893 |
| work_keys_str_mv | AT lachheba advancedintelligentcontrolforphotovoltaicvehicletogridintegration AT chroutaj advancedintelligentcontrolforphotovoltaicvehicletogridintegration AT telmoudiaj advancedintelligentcontrolforphotovoltaicvehicletogridintegration AT zaafouria advancedintelligentcontrolforphotovoltaicvehicletogridintegration |