Improved Identification of Nonlinear Dynamic Systems using Artificial Immune System

Abstract

Over the recent few years the area of artificial immune system (AIS) has drawn attention of many researchers due to its broad applicability to different fields. In this paper the AIS technique has been suitably applied to develop a new model for efficient identification of nonlinear dynamic system. Simulation study of few benchmark identification problems is carried out to show superior performance of the proposed model over the standard multilayer perceptron (MLP) approach in terms of response matching, number of training samples used and convergence speed achieved. Thus it is concluded that the AIS based model used is a preferred candidate for identification of nonlinear dynamic system.