Performance comparison of solar fed MPC and AI controller for fast battery charging in electric vehicles

Apoorva Srivastava
Department of Electrical Engineering, Babu Banarasi Das Institute of Technology and Management, Lucknow 226028, India
Mohammad Saif Raza
Department of Computer Science & Engineering, Babu Banarasi Das Institute of Technology and Management, Lucknow 226028, India
Faiz Haider
Department of Computer Science & Engineering, Maharishi University of Information Technology, Lucknow 226013, India
Prasant Shukla
Department of Computer Science & Engineering, Dr. A.P.J. Abdul Kalam Technical University, Lucknow 226031, India

Article ID: 4291

Keywords: solar PV-fed EV charging; model predictive control; deep reinforcement learning; fast battery charging; energy management; hardware-in-the-loop validation

Abstract

The rapid growth of electric vehicles (EVs) has increased the demand for charging infrastructure that is not only fast and efficient but also environmentally sustainable. Solar-powered EV charging stations, which integrate photovoltaic (PV) systems, offer a promising solution by reducing dependence on the electrical grid and lowering carbon emissions. However, the intermittent nature of solar energy creates significant challenges for maintaining stable and efficient fast-charging operations. This study evaluates the performance of different control strategies for a solar-powered EV fast-charging system. A comparative analysis was conducted between Model Predictive Control (MPC), Deep Reinforcement Learning (DRL), and Artificial Neural Network (ANN)-based controllers. The system consisted of a 100 kWp PV array, a 400 V DC bus, a bidirectional DC–DC converter operating at 20 kHz, and a 60 kWh EV battery charged at 1C–2C rates. The MPC controller was designed with a prediction horizon of 10, a control horizon of 3, and a sampling time of 100 μs using quadratic cost optimization, while the DRL controller employed a Deep Q-Network framework. Simulation results demonstrated that the DRL-based controller achieved superior performance under varying irradiance conditions. Compared with MPC, it increased solar energy utilization by 8%, improved charging efficiency by 12.8%, and reduced battery degradation by approximately 15% over 1000 charge–discharge cycles. In addition, DRL exhibited faster transient response, achieving system stabilization within 0.21 s during sudden irradiance changes, compared with 0.35 s for MPC. The findings indicate that advanced adaptive control strategies can enhance energy utilization, charging performance, and battery longevity in solar-powered EV charging applications.

Published

2026-06-04

How to Cite

Srivastava, A., Raza, M. S., Haider, F., & Shukla, P. (2026). Performance comparison of solar fed MPC and AI controller for fast battery charging in electric vehicles. Energy Storage and Conversion, 4(2). https://doi.org/10.59400/esc4291

Download Citation

Issue

Vol. 4 No. 2 (2026)

Section

Article

This work is licensed under a Creative Commons Attribution 4.0 International License.

References

[1]Hakam Y, Gaga A, Tabaa M, et al. Enhancing Electric Vehicle Charger Performance with Synchronous Boost and Model Predictive Control for Vehicle-to-Grid Integration. Energies. 2024; 17(7): 1787. doi: 10.3390/en17071787

[2]Luetz JM, Nichols E, Du Plessis K, et al. Spirituality and Sustainable Development: A Systematic Word Frequency Analysis and an Agenda for Research in Pacific Island Countries. Sustainability. 2023; 15(3): 2201. doi: 10.3390/su15032201

[3]Minchala-Ávila C, Arévalo P, Ochoa-Correa D. A Systematic Review of Model Predictive Control for Robust and Efficient Energy Management in Electric Vehicle Integration and V2G Applications. Modelling. 2025; 6(1): 20. Available online: https://www.mdpi.com/2673-3951/6/1/20

[4]Srivastava A, Yadav V, Yadav V et al. Performance comparison of PI and AI-based controllers for solar PV-fed fast electric vehicle battery charging systems. Energy Storage and Conversion. 2026; doi: 10.59400/esc4074

[5]Wen T, He J, Jiang L, et al. A simple and flexible bootstrap-based framework to quantify epistemic uncertainty of ground motion models by light gradient boosting machine. Applied Soft Computing. 2024; 152: 111195. doi: 10.1016/j.asoc.2023.111195

[6]Lin H, Cai C, Chen J, et al. Modulation and Control Independent Dead-Zone Compensation for H-Bridge Converters: A Simplified Digital Logic Scheme. IEEE Transactions on Industrial Electronics. 2024; 71(11): 15239–15244. doi: 10.1109/TIE.2024.3370975

[7]Hassan AM, Ababneh J, Attar H, et al. Reinforcement learning algorithm for improving speed response of a five-phase permanent magnet synchronous motor based model predictive control. PLOS ONE. 2025; 20(1): e0316326. doi: 10.1371/journal.pone.0316326

[8]Huang Z, Gong J, Xiao X, et al. Artificial Intelligence and Digital Twin Technologies for Power Converter Control in Transportation Applications: A Review. IET Power Electronics. 2025; 18(1): e70013. doi: 10.1049/pel2.70013

[9]Carvajal CP, Andaluz VH, Roberti F, et al. Path-following control for aerial manipulators robots with priority on energy saving. Control Engineering Practice. 2023; 131: 105401. doi: 10.1016/j.conengprac.2022.105401

[10]Aljundi K, Figueiredo A, Vieira A, et al. Geothermal energy system application: From basic standard performance to sustainability reflection. Renewable Energy. 2024; 220: 119612. doi: 10.1016/j.renene.2023.119612

[11]An Z, Zhao Y, Du X, et al. Experimental research on thermal-electrical behavior and mechanism during external short circuit for LiFePO4 Li-ion battery. Applied Energy. 2023; 332: 120519. doi: 10.1016/j.apenergy.2022.120519

[12]Zhang Y, Yang Q, An D, et al. Multistep Multiagent Reinforcement Learning for Optimal Energy Schedule Strategy of Charging Stations in Smart Grid. IEEE Transactions on Cybernetics. 2023; 53(7): 4292–4305. doi: 10.1109/TCYB.2022.3165074

[13]Fortuna C, Yetgin H, Mohorčič M. Smart Infrastructures: Artificial Intelligence-Enabled Lifecycle Automation. IEEE Industrial Electronics Magazine. 2023; 17(2): 37–47. doi: 10.1109/MIE.2022.3165673

[14]Golestan S, Golmohamadi H, Sinha R, et al. Real-Time Simulation and Hardware-in-the-Loop Testing Based on OPAL-RT ePHASORSIM: A Review of Recent Advances and a Simple Validation in EV Charging Management Systems. Energies. 2024; 17(19): 4893. doi: 10.3390/en17194893

[15]Zahraoui Y, Korõtko T, Rosin A, et al. Market Mechanisms and Trading in Microgrid Local Electricity Markets: A Comprehensive Review. Energies. 2023; 16(5): 2145. doi: 10.3390/en16052145

[16]Dehkordi NM, Nekoukar V. Adaptive distributed stochastic deep reinforcement learning control for voltage and frequency restoration in islanded AC microgrids with communication noise and delay. Scientific Reports. 2025; 15(1): 27315. doi: 10.1038/s41598-025-13010-6

[17]Nabih A, Li Q. Design of 98.8% Efficient 400-to-48-V LLC Converter With Optimized Matrix Transformer and Matrix Inductor. IEEE Transactions on Power Electronics. 2023; 38(6): 7207–7225. doi: 10.1109/TPEL.2023.3244869

[18]Gao B, Zhu Y, Li Y. Optimal Operation Strategy Analysis with Scenario Generation Method Based on Principal Component Analysis, Density Canopy, and K-medoids for Integrated Energy Systems. Journal of Modern Power Systems and Clean Energy. 2024; 12(1): 89–100. doi: 10.35833/MPCE.2022.000681

[19]Su T, Zhao J, Yao Y, et al. Safe Reinforcement Learning-Based Transient Stability Control for Islanded Microgrids With Topology Reconfiguration. IEEE Transactions on Smart Grid. 2025; 16(4): 3432–3444. doi: 10.1109/TSG.2025.3569696

[20]Dong J, Guo X, Zhang C, et al. Multi-level monitoring method based on slow independent component analysis-tensor decomposition for industrial batch processes. Measurement. 2025; 241: 115610. doi: 10.1016/j.measurement.2024.115610

[21]Zheng H, Du Q, Mo S, et al. Improved marine predator MPPT algorithm for photovoltaic systems in partial shading conditions. Scientific Reports. 2025; 15(1): 21092. doi: 10.1038/s41598-025-06408-9

[22]Li G, Wu J, Li S, et al. Multitentacle Federated Learning Over Software-Defined Industrial Internet of Things Against Adaptive Poisoning Attacks. IEEE Transactions on Industrial Informatics. 2023; 19(2): 1260–1269. doi: 10.1109/TII.2022.3173996

[23]Demeke W, Ryu B, Ryu S. Machine learning-based optimization of segmented thermoelectric power generators using temperature-dependent performance properties. Applied Energy. 2024; 355: 122216. doi: 10.1016/j.apenergy.2023.122216

[24]Nagadurga T, Raju VD, Barnawi AB, et al. Global MPPT optimization for partially shaded photovoltaic systems. Scientific Reports. 2025; 15(1): 10831. doi: 10.1038/s41598-025-89694-7

[25]Mollik Babu R, Alam MS, Islam A, et al. Toward Sustainable and Clean Energy Futures: A Techno-Economic Review of Solar PV Systems, Challenges, and Opportunities. IEEE Access. 2025; 13: 169720–169757. doi: 10.1109/ACCESS.2025.3614771

[26]Mishra A, Sahoo UK, Maiti S. Robust Structured Sparsity-Based Fused Lasso Estimator With Sensor Position Uncertainty. IEEE Transactions on Circuits and Systems II: Express Briefs. 2024; 71(4): 2449–2453. doi: 10.1109/TCSII.2023.3330151

[27]Santarelli C, Helbig C, Li A, et al. A Multi-Disciplinary Approach for the Electrical and Thermal Characterization of Battery Packs—Case Study for an Electric Race Car. World Electric Vehicle Journal. 2023; 14(4): 102. doi: 10.3390/wevj14040102

[28]Wu M, Ma D, Xiong K, et al. Optimizing load frequency control in isolated island city microgrids: a deep graph reinforcement learning approach with data enhancement across extensive scenarios. Frontiers in Energy Research. 2025; 12: 1384995. doi: 10.3389/fenrg.2024.1384995

[29]Zhang L, Ye H, Ding F, et al. Increasing PV Hosting Capacity With an Adjustable Hybrid Power Flow Model. IEEE Transactions on Sustainable Energy. 2023; 14(1): 409–422. doi: 10.1109/TSTE.2022.3215287

Editor-in-Chief

Prof. Xiaohu Yang

Xi'an Jiaotong University, China

eISSN

3029-2778

Publication Frequency

Quarterly

Indexing

Scopus

CAS

Scilit Database (Switzerland)

J-Gate

All the publications will be archived by the PKP Preservation Network for long-term electronic preservation.

Authors are encouraged to self-archive the final version of their published articles into institutional repositories (such as those listed in the Directory of Open Access Repositories).

Authors are also encouraged to use the final PDF version published on the website of Academic Publishing.

About the Publisher

Academic Publishing insists on taking academic exchange and publication as the main line, carrying out comprehensive management based on science and technology, and fully exploring excellent international publishing resources. Within 5 years, it will form a strategic framework and scale with science (S), technology (T), medicine (M), education (E), and humanities and arts (H) as the main publishing fields. Academic Publishing is headquartered in Singapore and based in Malaysia, with the United States and China providing the main scientific and academic resources. At the same time, it has established long-term good cooperative relations with other publishing companies, scientific research communities, and academic organizations in more than a dozen countries and regions. Academic Publishing uses English and Chinese as its main publishing languages, mainly publishing books, journals, and conference papers in print and online. The vast majority of publications follow the international open access policy, providing stable and long-term quality and professional publications. With the joint efforts of the expert team and our professional editorial team, our publications will gradually be indexed by international databases in stages to provide convenient and professional retrieval for various scholars. At the same time, manuscripts we accept will be subject to the peer review principle, and cutting-edge and innovative research articles will be preferentially accepted for peer reference and discussion. All kinds of our publications are welcome for peer to contribute, access, and download.

more

Volume Arrangement

2026

2025

2024

2023

Featured Articles

Nitrided Copper-Iron Composite Oxides Derived from Layered Double Hydroxides for Enhanced Carbon Dioxide Electroreduction to Methane and Formic Acid

The reduction of carbon dioxide to valuable chemical products is a promising solution to address carbon balance and energy issues. Herein, amorphous nitrided copper-iron oxides are prepared by gas-phase nitriding of Cu Fe-layered double hydroxide precursor s with urea as nitrogen source. Amorphous materials are more likely to generate defect vacancies during the reaction process, and these vacancies can function as active sites for catalytic reactions. Therefore, the obtained materials show high activity for CO 2 electroreduction to methane and formic acid, achieving a total Faraday efficiency of 74.7% at −0.7 V vs RHE and exhibiting a continuous 10 h durability in the H-cell. The uniformly distributed Cu + sites act as active sites by losing electrons to activate CO 2 . During the CO 2 electroreduction , CO 2 is converted to *COOH via proton-electron coupling, *COOH combines directly with a proton in solution to produce the HCOOH product, and the other part of *COOH undergoes a protonated dehydration process to form the *CHO intermediate which dehydrates again to form CH 4 . This study provides a new approach for designing CO 2 electroreduction catalysts.

AlZnO magnetron sputtered thin film for photovoltaic application

Aluminum zinc oxide (AZO) is a nontoxic and a low-cost material that finds application as a transparent conducting electrode in photovoltaic devices. In this study the (direct current) DC magnetron sputtering of AZO films is carried out at different deposition times of 5, 10, 15, 20 and 25 min’s at room temperature and it’s structural, optical, electrical and morphological properties are studied for its use as a front contact for thin film solar cell application. The structural study suggests that the preferred orientation of grains along (002) plane having hexagonal structure and the optical and the electrical studies suggest that the films show an average transmission of 70% and a resistivity of the order of 10 -4 Ωcm. On the other hand, the scanning electron microscopy (SEM) images suggest the formation of packed grains having a homogeneous surface. Moreover in order to study the optoelectronic properties of prepared samples, the electronic and optical calculations of the AZO are performed by the first-principles calculations using density functional theory (DFT).

Investigation of self-excited induction generator for supporting domestic loads and its extension to a microgrid

This study provides a technical and financial analysis for the incorporation of a microgrid structure with a wind energy conversion system for producing electricity. This study’s primary aim is to provide solutions for issues that arise when isolated induction generators are employed with microgrids. A closed-loop smart electronic load controller is used to regulate the loads in the system that are supplied by the generation. A switched variable capacitor bank is used to supply reactive power initially during a voltage dip at varying winds and loads to sustain the voltage profile. Additionally, a simple voltage control loop-based controller for the generator side converter maintains the voltage at a steady level. Using HOMER software, an economic study of the suggested wind-based microgrid structure is also presented. A laboratory experimental setup is used to support the MATLAB/Simulink study of the proposed method and its control. The findings back the feasibility of implementing the suggested plan in grid-isolated regions for supplying critical loads.

Research progress on ZnO/MoS₂/rGO ternary photocatalysts

Energy shortage and environmental pollution have become one of the important global issues, and semiconductor photocatalytic technology is considered as one of the effective means to solve these problems. As a new and efficient green material, ZnO has attracted wide attention. ZnO is widely used in the field of photocatalysis due to its non-toxicity, low cost, environmental friendliness, adjustable band gap, high electron density, and chemical stability. However, the recombination of photogenerated charge carriers in ZnO hinders its practical application and lowers the utilization efficiency of visible light. On the other hand, molybdenum disulfide/reduced graphene oxide (MoS₂ /rGO), as a binary non-precious metal co-catalyst, has a larger specific surface area, suitable band gap width, and visible light response capability compared to single-phase graphene co-catalyst. Therefore, introducing MoS₂ /rGO co-catalyst into the ZnO system can provide more active sites, reduce the probability of photogenerated charge carrier recombination, and improve the utilization efficiency of visible light. In this review, we summarize the hydrothermal synthesis methods for preparing this highly demanded nanocomposite material, including one-step and stepwise methods. Subsequently, we elaborate on the mechanism of enhancing light absorption and achieving efficient electron-hole separation behavior in the ternary system heterojunction structure during the photocatalytic process. Due to its significant advantages, this ternary system heterojunction structure has been widely applied in the field of photocatalysis, including applications such as pollutant degradation, sterilization, and water splitting.

Wind power forecasting technologies: A review

This study addresses the critical role of wind power forecasting in ensuring stable and reliable power system operations. Wind Power Forecasting is critical for the efficient operation of plant, time scheduling, and it’s balancing of power generation with grid integration systems. Due to its dependency on dynamic climatic conditions and associated factors, accurate wind power forecasting is challenging. The research delves into various aspects, including input data, input selection techniques, data pre-processing, and forecasting methods, with the aim of motivating researchers to design highly efficient online/offline models on weather-based data. The overarching goal is to enhance the reliability and stability of power systems while optimizing energy resource utilization. The analysis reveals that hybrid models offer more accurate results, highlighting their significance in the current era. This study investigates different Wind Power Forecasting (WPF) models from existing literature, focusing on input variables, time horizons, climatic conditions, pre-processing techniques, and sample sizes that affect model accuracy. It covers statistical models like ARMA and ARIMA, along with AI techniques including Deep Learning (DL), Machine Learning (ML), and neural networks, to estimate wind power.

fmos