神经网络训练参数等对控制效果的影响

Influence of Neural Network Training Parameter etc on the Control Effect

本文在简要分析基于神经网络控制优点的基础上，选用ＣＭＡＣ神经网络，设计相应的控制器，采用数值仿真方法，分别调整各项参数进行神经网络学习控制的实验研究，得到这些参数对系统控制效果的影响及其规律，这对基于神经网络控制器的设计、应用以及性能分析等工作具有指导意义．

On the basis of briefly analysing the advantages of Neural Network control, this paper usesthe CMAC Neural Network to design the corresponding controller. By the simulation, we have donesome experimental study on Neural Network learning control to adjust separately every parpmeter.Finally, we obtained the influence and law of these parameters to system control, they have guidingsignificance to the design,application,and performance analysis of Neural Network controller.