改进的神经网络自学习鲁棒自适应控制器及性能分析

介绍了神经网络自学习鲁棒白适应控制器，该控制器可以应用在模型未知的控制系统中。针对该控制器所存在的问题进行改进，选用具有自适应能力的PID替代固定增益比例的RC控制器，而且对控制方式进行调整，并对改进后的控制器进行了性能分析。分析结果说明，改进后的控制器鲁棒性、自适应更强。

Improved Neural Network Robust Adaptive Controller and Performance Analysis

Introduction was made to a neural network robust adaptive controller, which can be applied in model unknown control system. Improvement was made to the existing problems of the controller, by selecting PID with adaptive ability to replace RC controller with f ixed gain proportion, and control method was adjusted and analysis was also made to the performance of the controller. Results of analysis show that improved controller is stronger in robust adaptive.