DESIGN AND IMPLEMENTATION OF A THREE-PHASE ISOLATED MULTILEVEL AC–DC CONVERTER FOR DUAL EV CHARGING
Keywords:
Electric Vehicle Charging, Artificial Neural Network (ANN), Multilevel Converter, AC–DC Converter, Dual Battery Charging, Power Factor Correction.Abstract
As more EVs hit the road, the demand for efficient, reliable and smart charging infrastructure that can support the simultaneous charging of multiple vehicles is increasing. In this work, a Three-Phase Isolated Multilevel Converter with ANN Control for simultaneous charging of Dual Electric Vehicle Batteries is presented. A two-stage power conversion architecture is used, which includes a three-phase multilevel AC–DC converter and a dual-output isolated DC–DC converter. The AC–DC converter produces power factor correction and stabilizes a DC-link voltage and the isolated DC–DC converter enables two EV battery packs to be charged independently and simultaneously. An Artificial Neural Network (ANN)-based controller is developed to control the output voltage, boost the accuracy of power sharing and adjust to the different charging conditions to improve the system performance. The ANN controller always monitors the behavior of the system and provides the best possible control actions to minimize voltage fluctuations, reduce Total Harmonic Distortion (THD) and keep the input power factor high. Using multilevel converter technology, better output waveforms, lower switching stress on semiconductor devices and higher conversion efficiency can be attained. In addition, the isolated converter structure guarantees electrical safety and fault tolerance and provides flexibility in the operation of different battery ratings. The whole system is modeled and simulated in MATLAB/Simulink for various charging conditions to examine the dynamic conditions and system stability. The simulation results demonstrate the superior performance of the proposed ANN-based charging system in terms of efficiency, power quality, dual battery system load-balancing and reliability compared to the conventional charging system. Hence, the architecture proposed is a potential solution for the future smart EV charging station and advanced electric transportation infrastructure.
