Abstract: Since the multi-parameter identification of permanent magnet synchronous motor (PMSM) has slow speed and low accuracy, a parameter identification method based on the improved salp swarm algorithm was proposed in this paper. The algorithm firstly adopted the self-adaptive evaluation-move strategy and neighborhood optimum guide strategy based Von Neumann topology to update the position of followers twice, which strengthened information cooperation in the population and accelerated the convergence rate of parameter identification. Secondly, the algorithm used the opposition-based learning strategy to perturb the population position with a certain mutation probability, that avoided local optimum and misconvergence of the parameters. The simulation results show that this algorithm can identify PMSM parameter quickly and accurately.

Key words: salp swarm algorithm, Von Neumann topology, opposition-based learning, PMSM, multi-parameter identification