Brushless direct current (BLDC) motors have become the preferred choice for electric vehicle (EV) power trains due to their exceptional performance attributes, including fast dynamic response, high efficiency, durability, low acoustic noise, and minimized electromagnetic interference (EMI), which is critical for applications sensitive to electrical noise. BLDC motors achieve precise electronic commutation via an inverter and a rotor position sensor; however, to reduce costs, sensorless control methods that estimate rotor position without physical sensors are often employed. This study presents an advanced control strategy for sensorless BLDC motor speed regulation using a wavelet neural network (WNN) optimized with a fractional‐order proportional integral derivative (WNN‐FOPID) controller. WNNs are well‐suited for detecting intricate patterns in system data, enhancing control accuracy. The FOPID gains are optimally tuned using the random weighted chimp optimization (RW‐CHO) algorithm, an enhanced version of the classic chimp optimization algorithm (ChOA). Performance evaluations demonstrate that the proposed approach achieves rapid settling (0.29343 s) and response times (0.24013 s), a control signal peak of 6250 within 0.02 s, indicating the maximum control output generated by the proposed system. It also shows a superior convergence rate (1.21% improvement), minimal error statistics (−1.487), and enhanced stability (rise time of 0.29249 s; settling time of 0.3625 s). Compared with existing techniques, the proposed WNN‐FOPID‐RW‐ChOA model consistently achieves superior rotor speed control and precision, establishing a robust solution for sensorless BLDC motor applications in EVs.
Purushothaman et al. (Thu,) studied this question.