Neural network based predictive functional control for nonlinear systems
基于神经网络的非线性系统预测函数控制

A neural network predictive functional control method and its convergence analysis for discrete nonlinear systems are presented by using the local linearization of nonlinear activation functions. The method gives a direct and effective multi-step predicting method and uses linear methods to obtain the control law and avoids the complicated nonlinear optimization. Simulation results are presented in the article showing the efficiency of this method.