|
标题：基于神经网络的二自由度永磁同步电机控制 |
|
作者：张涛，余海涛，沈天骄，等 |
|
| 2016年第12期 访问次数：250次 |
|
摘要：永磁同步电机矢量控制的目标是使其具有良好的跟随性能和抗干扰性能。在转速环的控制中，采用传统的反馈控制不能同时保证良好的跟随性和抗干扰性；采用二自由度控制理论，两个控制器分别进行设计，设计两组控制参数，可以很好地实现控制目标。设计了前馈加反馈的二自由度控制器，反馈控制器用以抑制负载变化时的转速波动；前馈控制器用以实现转速的快速跟踪。针对该电机参数的时变性和不确定性，设计的前馈控制器采用神经网络控制，对该电机逆模型进行识别，实现转速快速无静差跟踪给定转速，同时提高了整个控制系统的鲁棒性。仿真结果验证了，基于神经网络逆识别的二自由度控制具有很好的动态性能、静态性能和良好的鲁棒性。
关键词：永磁同步电机；二自由度控制；前馈控制；神经网络逆识别
Abstract: The vector control goal of Permanent magnet synchronous motor is to make it good tracking performance and disturbance rejection. In the speed loop control, it cannot guarantee good following performance and disturbance rejection by conventional feedback control at the same time; with two degree of freedom control, using two controllers, one is designed for tracking, the other is designed for disturbance rejection. This paper used two degree of freedom controller of feedforward control and feedback control. Feedback controller restrained the fluctuation of speed when the load changed. Feedforward controller kept fast tracking. As for permanent magnet synchronous motor (PMSM), it has timevarying parametric and uncertainty. By the neural network control of feedforward controller, realized the identification of the inverse model of the PMSM. Model was made in Matlab/Simulink,results show the effectiveness and feasibility of the theory.
Key words: permanent magnet synchronous motor；twodegreeoffreedom theory; feedforward control; neural network inverse recognition
|
|
| |
|
