AI-Based Intelligent Control Algorithm for Solar PV Powered Bidirectional DC–DC Converter in Electric Vehicle Battery Charging Systems
TL;DR
The proposed intelligent charging architecture offers a practical solution for next-generation renewable-energy-based EV charging stations and supports future smart-grid and vehicle-to-grid applications.
Abstract
The rapid adoption of electric vehicles (EVs) has significantly increased the demand for efficient, intelligent, and sustainable charging infrastructures. Conventional charging systems primarily depend on utility-grid power, resulting in increased operational costs, carbon emissions, and power quality challenges during peak demand periods. Integrating photovoltaic (PV) generation with bidirectional DC–DC converters provide a promising solution by enabling renewable-energy-powered charging and vehicle-to-grid (V2G) operation. However, the intermittent nature of solar irradiance and varying battery operating conditions require advanced control strategies beyond conventional proportional–integral (PI) controllers.This paper presents an Artificial Intelligence (AI)-based intelligent control algorithm for a solar photovoltaic powered bidirectional DC–DC converter used in electric vehicle battery charging systems. The proposed controller combines AI-assisted maximum power point tracking (MPPT), adaptive battery charging, and converter duty-cycle optimization to improve energy conversion efficiency, charging performance, and system stability under dynamically changing environmental conditions.A mathematical model of the PV array, lithium-ion battery, and bidirectional DC–DC converter is developed to describe the system behaviour. The AI controller continuously predicts the optimum converter operating point using real-time measurements of irradiance, temperature, battery state of charge (SOC), battery voltage, and load demand. The proposed strategy is evaluated using MATLAB/Simulink under varying irradiance levels, battery SOC, and charging conditions. Simulation results demonstrate superior MPPT accuracy, reduced charging time, lower converter losses, improved DC bus voltage regulation, and higher overall efficiency compared with conventional PI and fuzzy logic controllers.The proposed intelligent charging architecture offers a practical solution for next-generation renewable-energy-based EV charging stations and supports future smart-grid and vehicle-to-grid applications.