【摘要】 为改善永磁同步电机（PMSM）闭环系统调速性能，使其更好满足快速性与稳定性的要求，提出一种变置信度和自适应学习率优化小脑模型神经网络（MC-CMAC）的永磁同步电机调速控制方法。首先，引入置信度因子θ改进权值调整方式，避免权值修正的平均分配；在此基础上，增加平衡学习系数σ，进一步优化网络学习结构；最后，提出一种网络学习率自适应调整方法，在应对复杂工况时仍能保持良好的学习增益。搭建永磁同步电机矢量控制模型，仿真结果表明，与PI、常规CMAC、无模型滑模控制相比，此方法响应速度更快，动态性能更好。更多还原

【Abstract】 To improve the speed control performance of the closed-loop system of permanent magnet synchronous motor(PMSM) to better meet the requirements of speed and stability, a control method for permanent magnet synchronous motor(PMSM) is proposed, in which a variable confidence and adaptive learning rate optimises the cerebellar model neural network(MC-CMAC). Firstly, a confidence factor is introduced to improve the weight adjustment method to avoid the average distribution of the weight correction; on this basis, the balanced learning coefficient is added to further optimise the network learning structure; finally, an adaptive adjustment method of the network learning rate is proposed, which can still maintain good learning gain while dealing with complex working conditions. The vector control model of permanent magnet synchronous motor is constructed, and the simulation results show that the method in this paper possesses a faster response speed and better dynamic performance compared with that of PI, conventional CMAC, and modelless sliding mode control.更多还原