Generalized Regression Neural Network Based Predictive Model of Nonlinear System

Abstract

Generalized Regression Neural Network (GRNN) is usually applied to the Function approximation. This paper, based on the principle of GRNN, presents a method for the predictive model of nonlinear complex system. The presented algorithm is applied to the learning and predicting process for the system modeling. The simulations show the described method has good effects on predicting the dynamic process of the nonlinear model, and could be applied on the predictive control for nonlinear systems satisfactorily. DOI : http://dx.doi.org/10.11591/telkomnika.v12i3.4560 Full Text: PDF