A Neural Network Decoupling Strategy for a Class of Nonlinear Discrete Time Systems
一类非线性离散时间系统的神经网络解耦策略

A neural network is considered to be used as a compensator for input output decoupling of a class of nonlinear discrete time systems. A necessary and sufficient condition for the solvability of the decoupling problem for the class of discrete time systems is given. It is also shown that if the decouling problem is solvable, the modified systems can be linear and the poles of the modified systems can be freely assigned. Based on this result, a strategy for realizing decoupling via neural networks is proposed. Simulation results supports our theory and the decoupling strategy proposed in this paper.