摘要：工程实际中，考虑到伺服系统高精度、动静态特性好的要求，其控制参数应根据负载转动惯量的变化而作相应的调整。因此，辨识负载的转动惯量成为新的研究热点。采用递推最小二乘法设计了永磁同步电机的负载转动惯量辨识算法，并采用平均值算法、变遗忘因子来优化该算法。仿真结果表明了该方法收敛时间由017 s减少到005 s，辨识误差由82%降至57%，辨识曲线平滑度高，具有更好的辨识特性。
关键词：永磁同步电机；最小二乘法；变遗忘因子
Abstract: In practical project, the parameters of control should make adjustmentswhen rotational inertia was changed consider of highprecision and the static and dynamic characteristics of servo system. Therefore, the identification of the load inertia has become a hot research. An identification method for the inertia of load in permanent magnet synchronous motor(PMSM) was designed based on recursive least square(RLS), which was optimized by algorithm of average and alter forgetting factor. Simulation results show that the convergence time of the system reduce from 017 s to 005 s, the error lessen from 82% to 57%. The character of the improved algorithm is better as the identification curve owes a high smoothness.
Key words: PMSM;recursive least square(RLS);alter forgetting factor