Direct Adaptive Decoupling Control of Nonlinear Systems Based on Neural Networks and Multiple Models

A direct adaptive decoupling controller is presented for a class of uncertain nonlinear multivariable discrete time dynamical systems. The direct adaptive decoupling controller is composed of a linear direct adaptive decoupling controller, a neural network nonlinear direct adaptive decoupling controller and a switching mechanism. The linear decoupling controller can provide boundedness of the input and output signals, and the nonlinear decoupling controller can improve performance of the system, while the switching mechanism is utilized to obtain the improved system performance and stability simultaneously. Theory analysis and simulation results are presented to show the effectiveness of the proposed method.