Online Decoupling and Control of Multivariable Nonlinear System Based on Neural Network

Aimed at neural network can approach any nonlinear dynamic system with arbitrary accuracy, the frame of distributed NN decoupling system are proposed to decouple the MIMO nonlinear system. The learning algorithm of NN is online, which makes a set of cross-correlation function as the target function, and adopts hybrid genetic algorithm to train neural decoupler. When the decoupling is completed, the single nerve cell self-adaptor PID is used to control two SISO systems. The efficiency of the algorithm has been shown by numerical simulations combining nonlinear system