关键词: 永磁同步电机, 参数辨识, 无差拍电流预测, 模型参考自适应, 粒子群算法

Abstract: Permanent magnet synchronous motor ( PMSM) is a strongly nonlinear system in practical application. The parameters of PMSM will change due to temperature and magnetic saturation, which can affect the control effect of PMSM. An online parameter identification algorithm was proposed based on hybrid optimization of model reference adaptive system ( MRAS) and genetic particle swarm optimization ( GAPSO) . The electrical parameters of PMSM were initially identified
by MRAS, and the identification results provided direction for particle swarm optimization. Subsequently, the crossover mutation mechanism of the genetic algorithm was improved, and the preliminary identification results were introduced into the genetic particle swarm optimization algorithm for further optimization. The simulation results show that all parameters identified by the MRAS-GAPSO algorithm achieve high precision identification within 50 iterations, with a relative error of less than 1%. The simulation results also demonstrate the algorithm’ s high convergence performance and its ability to accurately identify parameters in real-time.

Key words: permanent magnet synchronous motor ( PMSM ), parameter identification, deadbeat predictive current control( DPCC), model reference adaptive system( MRAS), particle swarm optimization( PSO), genetic algorithm( GA)