基于神经网络和神经元PID的交流矢量自适应控制

AC vector control system based on neuron networks and neuron PID

磁链开环的异步电动机矢量控制系统尽管结构简单,但它的性能受转子电阻影响很大.为了解决这一问题,根据感应电机矢量控制原理,将单神经元自适应控制器和神经网络应用于磁链开环的交流矢量控制系统中,并用单神经元PID代替传统矢量控制中的转速调节器,用神经网络代替系统中的电流模型,利用神经元PID的自适应和神经网络的在线学习能力,使系统能够克服其运行中由于自身参数变化带来的影响,具有很强的鲁棒性.最后,应用MATLAB软件进行了仿真实验.结果表明,该系统不但结构简单,而且具有良好的动、静态性能.

The construction of asynchronous motor?drive vector control system with the open?loop magnetic flux linkage is simple,but its performance is badly affected by the rotor resistance.In order to solve this problem,first,according to the principle of the vector control of induction motors,a single neuron self?adaptive controller and nueral network are adopted in the AC vector control system with the open?loop magnetic flux linkage.Then,the speed regulator in the traditional vector control system is replaced by the single neuron PID,and the current mode in it is also replaced by the nueral network becauce of the self?adaptive ability of single neuron PID,the studied ability of nueral network online,the effects which are brought by the changes of system parameters during the process can be overcomed,and the system has good robust.Finally,the system is simulated by the MATLAB.The simulation result shows that this system is simple on its construction and has a better dynamic and static performance.