PMSM Parameter Identification Based on Chaotic Adaptive Search Grey Wolf Optimization Algorithm

Aiming at the problems of poor population diversity, slow speed of late identification, and low identification accuracy of traditional grey wolf algorithm (GWO), a chaotic adaptive search grey wolf optimization algorithm (CASGWO) for parameter identification of permanent magnet synchronous motor is proposed in this paper. Firstly, multiple low-dimensional chaotic mappings are combined; a composite chaotic system Tent-Logistic-Cosine is obtained; uniform populations are generated. So the population diversity and global search capability are improved. Then a segmented nonlinear search method is proposed, where the nonlinear decay factor quickly converges to the vicinity of the optimal solution in the first segment and slows down the convergence rate for local search in the second segment. Thus, the convergence rate is accelerated while the local search capability is enhanced. Finally, the adaptive inertia weights are adjusted according to the fitness values of different wolf pack iterations, and ω wolves approach the leader wolf pack with smaller fitness values at a faster speed. Therefore, the speed of search is again improved, and the local search ability of the algorithm is again enhanced. Experiments show that when identifying multiple parameters of resistance, inductance, and permanent magnet flux of a permanent magnet synchronous motor, the CASGWO method has good global and local search capability, with faster identification speed and higher identification accuracy than the traditional grey wolf algorithm. © 2024, Electromagnetics Academy. All rights reserved.