A Solar-Powered EV Charging Station with Neural Network

Abstract

The system for the charging electric vehicles (EVs) using solar photo--voltaic (PV) power, enhanced by a neural network. Solar energy is highlighted as a clean, renewable, and limitless resource that produces no greenhouse gas emissions. A key challenge with solar power is its variable output, especially on cloudy days. To overcome this, the paper Details with using an Adaptive Neuro-Fuzzy Inference System, ANFIS with Max Power Point Tracking, MPPT. This neural network-based technique optimizes the power output of solar panels, ensuring sufficient electricity generation for EV charging, even when sunlight is limited. The increasing adoption of EVs has spurred demand for more charging infrastructure. However, despite EVs themselves being emission-free, their charging often relies on conventional energy sources, which do impact the environment. This issue by proposing solar PV-powered EV charging stations. To ensure continuous power supply without adding strain to the main grid, these stations incorporate a battery stored system and grid support. The proposed system, which integrates the Adaptive Neuro--Fuzzy inference system with the neural network techniques, has been modelled and evaluated using Simulink

Introduction

The increasing demand for transportation has intensified global warming, largely due to greenhouse gas emissions from internal combustion engine vehicles. Electric Vehicles (EVs) offer a cleaner, more efficient alternative, but large-scale EV adoption creates challenges for the existing power grid because charging demand continues to rise. Conventional grid-powered EV charging also reduces environmental benefits, highlighting the need for renewable-energy-based charging systems.

To address this, the study proposes a solar-powered EV charging station integrated with a Battery Energy Storage System (BESS) and supported by the AC grid. The system uses neural-network-based power management and an ANFIS-based MPPT controller to ensure maximum power extraction from the solar PV array under changing temperature and irradiance conditions. When solar energy is high, PV power charges the EV and BESS; at night or low irradiance, the BESS supplies energy; the grid is used only when both PV and BESS cannot meet the load demand.

Methodology Overview

The system is modeled in MATLAB/Simulink with the following components:

• PV Array: 2500 W, connected to a 400 V DC bus through a boost converter.

• BESS: 240 V, 40 Ah lithium-ion battery, managed via a bidirectional converter.

• EV Battery: 240 V, 7 Ah lithium-ion battery starting at 9% SOC.

• AC Grid: 230 V, 50 Hz supply used as backup.

Control Strategies

• ANFIS-Based MPPT:
  Ensures optimal tracking of maximum PV power using fuzzy logic and neural network learning capabilities. It performs better than traditional MPPT algorithms (P&O, IncCond) by handling rapid environmental changes with higher precision and reduced oscillations.

• Neural Network Grid Controller:
  Determines the grid reference current based on PV output and BESS SOC to maintain stable system operation.

• BESS Control:
  Allows charging/discharging while ensuring SOC does not fall below 20%.

Power Management Logic

1. PV as the primary source.

2. Excess PV energy charges the BESS or feeds the grid.

3. BESS supplies energy when PV is insufficient.

4. Grid compensates only when both PV and BESS cannot meet demand.

Simulation Results

System performance is evaluated under five operating modes:

1. PV-only EV charging with constant temperature and variable irradiance.

2. PV charges both EV and BESS.

3. EV charging supported by both PV and BESS.

4. Night mode: BESS + grid supply power.

5. Full hybrid mode: PV, BESS, and grid work together with surplus fed into the grid.

Simulation graphs indicate:

• Stable PV voltage and current under MPPT control

• Efficient grid synchronization

• Smooth BESS charge/discharge with SOC maintenance

• Lower Total Harmonic Distortion (THD) using ANN/ANFIS control compared to PI control

Conclusion

The proposed PV-powered EV charging station, enhanced with neural network–based grid control and ANFIS-driven MPPT, ensures stable DC bus voltage and intelligent power flow between solar PV, BESS, and the grid it demonstrates reliable voltage regulation and efficient energy coordination Simulations verify its ability to maximize renewable utilization, maintain uninterrupted EV charging, and adapt to varying conditions, making it a scalable solution for future high-demand charging infrastructures.

References

[1] X. Sun, Z. Li, X. Wang, C. Li, “Technology Development of Electric Vehicles” A Review. Energies 2020, 13, 90. [2] M.A.H. Rafi, J.A. Bauman, “Comprehensive review of DC fast charging stations with energy storage: Architectures, power converters, and analysis” IEEE Trans. Transp. Electric. 2021, 7 345-368. [3] A. Hamidi, L. et. al \"EV Charging Station Integrating Renewable Energy and Second - Life Battery,\"2013 Int. Conf. Renew. Energy Res. Appl., no. October pp. 20--23, 2013.J. Clerk Maxwell, A Treatise on Electricity and Magnetism, 3rd ed., vol. 2. Oxford: Clarendon, 1892,pp. 68-73. [4] S. Akshya, A. Ravindran, A. S. Srinidhi S. Panda, and A. G. Kumar, “Grid integration for electric vehicle and photovoltaic panel for a smart home,” Proc. IEEE Int. Conf. Circuit Power Comput. Technol. ICCPCT 2017, pp. 1–8, 2017 doi: 10.1109/ICCPCT.2017.8074358. [5] A. S. Sener, “Improving the life-cycle and SOC of the battery of a modular electric vehicle using ultra-capacitor,” 8th Int Conf. Renew.Energy Res.Appl. ICRERA 2019, pp. 611–614, 2019, doi:10.1109/ICRERA47325.2019.8996616

Copyright

Copyright © 2025 Madipelli Abhinava, Dr. M. Sushama. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.