Development of Supervisory Self-Learning-Based Energy Management Controller to Control the Torque Ripples of Brushless DC Motor in Electric Vehicles Applications

The BLDC motor is becoming more popular in electric vehicles than other motors because of its high torque, robust control, etc. Antithetically, it is more challenging to regulate the torque ripples of a BLDC motor under various real-time conditions. To overcome these shortcomings, this study uses novel methodology to analyse the effectiveness of BLDC motor with different energy management controllers. In this study, to control the torque variations of a BLDC motor, supervisory, hybrid, intelligent, and PID controllers are designed using MATLAB/Simulink. This study integrates experimental and modelling techniques to understand the behaviour of a BLDC motor with various EMCs in real-time conditions. The simulation finding indicates that the supervisory controller exhibits a minimal settling time (0.05 s), rise time (0.01 s), and overshoot (0.93%) than other conventional controllers. It integrates knowledge of real-time rules and numerical data with a neural network in order to forecast and reduce error statistics in the BLDC motor. As well, the proposed supervisory approach significantly reduces the torque fluctuations of BLDC motors under various real-time operating conditions. Further, an experimental study is conducted to validate the results of the simulation; this work entails the development of efficiency maps for a variety of energy management controllers. The efficiency of the various controllers is 97, 91, 85, and 78%, respectively. The recommended supervisory strategy surpasses other controllers in terms of time-domain characteristics under different conditions. Therefore, the proposed controller will enhance the efficiency of the BLDC motor and extend driving range of EVs under various real-time driving situations.