Neural Network Based Predictive Control for Nonlinear Systems

A neural network based predictive controller design algorithm is introduced for nonlinear control systems. It is shown that the use of nonlinear programming techniques can be avoided by using a set of affine nonlinear predictors to predict the output of the nonlinear process. The new predictive controller, based on this design, is both simple and easy to implement in practice. An on-line weight learning algorithm based on neural networks is introduced and convergence of both the weights and estimation errors is established. Predictive controller design, based on the new procedure, is illustrated using a growing network example.