交流传动神经网络自适应规划控制系统的设计与仿真

The Design and Simulation of Neural Network Adaptive Programming Control for AC Drive System

将 Hopfield神经网络应用于交流传动系统的自适应控制 ,通过神经网络来规划交流调速系统的速度控制器动态输出 ,使速度控制器对某些参数变化具有良好的鲁棒性 .对于不可控的负载转矩 ,加入参数自动跟踪神经网络 ,构成具有参数在线跟踪功能的双神经网络自适应规划控制模式 ,进一步提高系统的性能 .仿真结果表明 。

A Hopfield neural network is used as an adaptive control of the AC drive system. The dynamic output of the speed controller is programmed through neural network, so that the speed controller has a good robustness against the variation of parameters. By adding the parameter auto tracking neural networks, a double neural network adaptive programming control model with the function of on line parameter tracking is constructed to overcome the uncontrollable load torque. Simulation results show that this system is capable of good dynamic performance.