Identification of nonlinear system using computational paradigms

System identification is the powerful technique to identify the mathematical model of the unknown system based on input-output data. It is difficult with ordinary methods when the system is nonlinear with uncertainty. The computational paradigm has very high & efficient self-learning ability. It has great potentialities for mapping of nonlinear system particularly for identification. ARX based neural network is designed with back propagation and genetic algorithm training algorithms for nonlinear system. The comparative analyses of these algorithms are discussed in this paper.