A New Approach to Control the On Board Charger in Electric Vehicles Using Convolutional Neural Networks

The electric vehicle is the area where advanced power electronic systems are maintained for the main purpose to improve efficient and reliable operation. The charging of batteries inside the vehicle can be affected by several criterions such as nonlinear power behavior and grid variations. In this paper, it is proposed to design and implement experimentally a novel control strategy for the dual active bridge converter within the on-board charger by using convolutional neural networks. The purpose of this study is to keep the switching duty cycle of the buck-boost converter and the phase-shifting angle at optimal values in order to enhance overall system efficiency. The effectiveness of the proposed controller is evaluated through numerical examples and with an experimental hardware prototype. In order to feature more the efficiency of the proposed approach implemented, a comparative study has been conducted between the results achieved by the proposed method and that of a traditional Proportional-Integral-Derivative (PID) feedback controller.