Development of an AI-Driven Pulse Charging System with Real-Time Battery Health Monitoring for EVS

Abstract

Electric Vehicles (EVs) have emerged as a sustainable alternative to conventional transportation, necessitating advanced energy management strategies for efficient and reliable battery utilization. This paper presents the modeling, simulation, and performance evaluation of an intelligent Battery Management System (BMS) integrated with a multilevel power converter, controlled using an Artificial Neural Network (ANN). The proposed ANN-based controller is trained to estimate the State of Charge (SoC) of individual battery modules and to make real-time intelligent switching decisions for optimal energy management. The multilevel converter enables bidirectional power flow, supporting both charging and discharging operations between the battery pack, load, and grid. A modular battery bank consisting of multiple parallel-connected batteries is managed using ANN-driven logic to achieve SoC balancing, over-discharge prevention, and improved overall system efficiency. The complete system is developed and simulated in MATLAB/Simulink and validated under dynamic operating conditions. Simulation results demonstrate accurate SoC estimation, stable voltage regulation, effective battery selection, and robust system performance. The proposed approach contributes toward the development of intelligent and reliable EV energy management and charging infrastructure, with enhanced battery longevity and improved operational efficiency.