Abstract
This paper presents MIMO predictive control using recurrent neural networks for a class of nonlinear discrete-time systems. The recurrent-neural-network-based predictive control law is developed from the optimization of a generalized predictive performance criterion. A real-time adaptive control algorithm, including a neural predictor and a neural predictive controller, is proposed; the learning rates for both the neural predictor and controller are determined based on Lyapunov stability theory. Simulation results reveals that the proposed control strategy gives satisfactory tracking and disturbance rejection performance for two illustrative nonlinear multivariable systems.

This publication has 23 references indexed in Scilit: