ANN Controlled Bidirectional Totem Pole AC-DC Converter with Isolated DAB for High-Efficiency EV Charging

Abstract

This paper presents the design and control of an ANN-based bidirectional Totem-Pole AC/DC converter integrated with an isolated Dual Active Bridge (DAB) DC-DC converter for high-efficiency electric vehicle (EV) charging applications. The proposed system enables efficient bidirectional power transfer between the AC grid and EV battery through a common DC-link structure. An Artificial Neural Network (ANN) controller is implemented to regulate the DC-link voltage, improve dynamic response, and control power flow under varying operating conditions. The Totem-Pole converter enhances power factor correction and reduces total harmonic distortion by replacing conventional diode bridge rectifiers with high-frequency active switches. The isolated DAB converter provides efficient bidirectional energy transfer and electrical isolation between the DC-link and EV battery. The ANN controller dynamically adjusts the duty cycle and phase shift angle for precise voltage and current regulation. Simulation results under varying load conditions demonstrate improved DC-link voltage stability, reduced harmonic distortion, fast transient response, and enhanced system efficiency compared to conventional PI-controlled systems. The proposed system is highly suitable for fast EV charging stations, smart grid applications, and advanced power electronic systems.

Keywords

Bidirectional Converter, Totem-Pole Converter, Dual Active Bridge (DAB), Artificial Neural Network (ANN), Electric Vehicle Charging, Power Factor Correction (PFC), Total Harmonic Distortion (THD), DC-Link Voltage, MATLAB/Simulink, Smart Grid.