Artificial Intelligence-Driven Inductor Current Prediction in HD-SPWM-Controlled 3-Level DAB Converter for Fast EV Charging Application

The rapid proliferation of electric vehicles (EVs) has led to the emergence of the need for efficient and fast charging systems. This paper addresses the development of an artificial intelligence (AI) based model to predict the optimized peak inductor current in a high-definition sinusoidal pulse width modulation (HD-SPWM) controlled 3-level dual active bridge (DAB) converter for a fast charging electric vehicle application. Although the DAB converter topology has numerous advantages, which can be used for various power distribution systems due to the bidirectional power flow and high-power density, the DAB converter's performance level is significantly dependent on the performance of the model since the inductor current should accurately be predicted. The proposed AI-based model, with a novel long short-term memory (LSTM) approach, can predict the inductor current accurately based on the anticipated input parameters. The MATLAB/SIMULINK simulation platform has been used to train the model, indicating that the model can be robust under a wide arrangement of operating conditions. The error-free inductor current prediction would significantly affect the controller with excellent precision, minimal losses, and high-speed charging of the electric vehicle. A series of simulation tests has been executed to expose the promiscuity of the proposed AI-based model with conventional line-up.