Artificial Intelligence-Enabled Intelligent Monitoring for Anomaly Detection in Inductor Current of PSFB Converter for Fast EV Charging Application

On the whole, the number of electric vehicles (EVs) has been rising quickly worldwide. Meeting the high-performance power-conversion demands to facilitate a more sustainable future requires isolated DC/DC converters like: phase shifted full bridge (PSFB) DC/DC converter. This article presents the development of artificial intelligence (AI)-based inductor current monitoring and anomaly detection in a pulse-width modulation (PWM) controlled phase shifted 3-level PSFB converter to be adopted for fast EV charger applications. A novel model has been proposed with a new chronological long short-term memory (LSTM) based autoencoder (AE) or LSTM-AE model which results in detection and diagnosis of such anomalies. MATLAB/Simulink is utilized for the validation of fast charging analysis and in the anomaly identification methodology, Python has been used as a programming language.