Multi-Objective Optimization of Permanent Magnet Synchronous Motor Based on Elite Retention Hybrid Simulated Annealing Algorithm

Aiming at the features of the multi-objective, multi-constraint and nonlinear parameter optimization of Permanent Magnet Synchronous Motor (PMSM), a modified hybrid simulated annealing genetic algorithm (SAGA) is adopted in this paper. Crossover mutation operator from genetic algorithm is introduced into simulated annealing process, and an elite reservation strategy is added into genetic selection operator to prevent losing excellent individuals. SAGA not only has the local search ability of the simulated annealing algorithm, but also has the ability of global optimization of genetic algorithm, which improves the premature convergence of the algorithm and the convergence speed. A multi-objective optimization model of structural parameters is established for a 4-pole 1.1 kW PMSM. The optimization algorithm and electromagnetic calculation program are combined to form a multi-objective optimization algorithm. The optimization results simulated by Ansoft show that the algorithm can get better results after optimization and improve the starting performance and operating characteristics of the motor.