Metaheuristic-Based Observer Design for Electromagnetic and Load Torque Estimation in PMSMs: A Comprehensive Review
Keywords:
PMSM, LTOs, Metaheuristic Optimization, FOC, MPCAbstract
Permanent Magnet Synchronous Motors (PMSMs) are widely used in electric vehicles and robotics due to their high efficiency and torque density. However, accurate torque and speed estimation without mechanical sensors remains a major challenge. To address this, observer-based techniques such as EKF, UKF, MRAS, SMO, and Luenberger observers have been extensively applied, along with adaptive, nonlinear, and AI-based observers.
Recent research trends focus on load torque observers for their superior disturbance rejection under dynamic conditions. Moreover, metaheuristic optimization algorithms such as PSO, GA, and GWO have been utilized to auto-tune observer and controller parameters, enhancing estimation accuracy and dynamic performance.
Modern control strategies like Field-Oriented Control (FOC) and Model Predictive Control (MPC) benefit from these optimized estimates, achieving higher efficiency and reduced torque ripple. Despite these advances, most studies rely on simulation, highlighting the need for real-time implementation and hardware validation, particularly in electric vehicle drive systems.