This study presents the design and performance evaluation of a bidirectional electric vehicle charging system integrating solar photovoltaic energy with an Artificial Neural Network based control strategy. The proposed architecture employs a modified Single-Ended Primary Inductor Converter capable of supporting both Grid-to-Vehicle and Vehicle-to-Home operating modes while maintaining stable bidirectional power flow between the grid, photovoltaic source, and EV battery. The ANN controller dynamically regulates the duty cycle of the MOSFET switches using battery current feedback and reference current signals, enabling adaptive control under varying solar irradiance and grid conditions. Simulation results indicate that the proposed system achieves charging efficiencies above 90% while maintaining stable operation for both 72 V and 240 V EV battery configurations. Compared with conventional proportional–integral control approaches, the ANN controller demonstrates faster transient response and improved current regulation during dynamic operating conditions. The integration of solar photovoltaic energy further reduces reliance on grid power and enhances renewable energy utilization in EV charging infrastructure. These results indicate that the proposed ANN-controlled bidirectional charging system provides an efficient and flexible solution for renewable-integrated EV charging applications.
Poojary et al. (Thu,) studied this question.