一类非线性系统迭代优化神经网络控制器

A Neural Network Based Iterative Optimal Controller for Nonlinear Systems

提出了一类非线性系统迭代优化神经网络控制器的设计方法．该方法是在正向神经网络辨识模型的基础上，应用迭代学习方法进行控制器设计．为了补偿辨识和迭代学习误差，给出了通过引入反馈补偿控制器提高控制精度的方法，仿真结果证明了该方法的有效性．

A neural network based iterative optimal controller for nonlinear systems is presented. The controller is designed based on the neural networks using iterative optimal methods. In order to reduce the errors of identification and iterative learning,a compensation controller is added in the control system. The simulation results show the efficiency of the proposed method.