Modeling and Implementation of Brain-Inspired Intelligent Controller for Permanent Magnet Synchronous Motor Drive

This paper presents a novel bio-inspired intelligent controller named the cerebro-cerebellar hybrid intelligent controller (CCIC) to significantly improve the dynamic performance of a nonlinear system, specifically a permanent magnet synchronous machine (PMSM). Conventional PMSM controllers often struggle to adapt to dynamic operating conditions and disturbances, leading to performance degradation. The CCIC was designed to address this challenge by drawing inspiration from the human brain's structure and function. The CCIC incorporates the functionalities of the cerebrum, cerebellum, and brainstem into a unified control system. The cerebrum component employs an emotional learning mechanism, enabling real-time adaptation based on operating conditions. The cerebellum component precisely regulates the PMSM's motion, ensuring smooth operation and minimizing tracking errors. The brainstem component ensures overall stability and robustness against external disturbances. To assess the efficacy of the proposed CCIC, its performance was evaluated on a PMSM drive under various operating scenarios, such as constant speed and torque, variable speed with constant torque, and variable torque with constant speed. To conclusively demonstrate the superiority of the CCIC, a comparative analysis was conducted against a state-of-the-art developed cerebellum controller and a conventional proportional-integral controller. The simulation results convincingly revealed that the CCIC outperformed both controllers in several critical metrics. This approach achieved demonstrably faster settling times, reduced rise times, minimized overshoot, and mitigated initial transient torque dips. Most importantly, the CCIC exhibited superior robustness against system disturbances and uncertainties, making it a highly reliable control solution for PMSMs.