NURAL NETWORK CONTROLLED THREE-PHASE MULTILEVEL ISOLATED CONVERTER

Abstract

This paper proposes an intelligent electric vehicle (EV) charging system based on a two-stage converter architecture capable of simultaneously charging dual EV batteries. The system incorporates a three-phase multilevel boost power factor correction (PFC) AC–DC converter in the front stage and a bidirectional isolated dual-output DC–DC converter with zero-voltage switching (ZVS) in the second stage. The independent operation of both stages ensures effective control without mutual interaction while maintaining stable performance in continuous conduction mode. To improve system efficiency and dynamic response, an Artificial Neural Network (ANN)-based controller is employed in place of conventional control methods. The ANN controller is trained to regulate output voltage, reduce total harmonic distortion (THD), and maintain a high power factor under varying grid and load conditions. Furthermore, it enables balanced power sharing between the two batteries, enhances transient response, and improves overall system stability. The proposed architecture achieves efficient energy conversion with reduced switching losses and reliable dual-battery management. Simulation studies conducted using MATLAB/Simulink validate the effectiveness of the system in terms of high efficiency, enhanced power quality, and intelligent control capability. Hence, the proposed ANN-based multilevel isolated converter offers a promising solution for advanced fast and smart EV charging applications.