On-line identification of continuous-time nonlinear systems using radial basis function networks and immune algorithm

This paper presents an on-line identification method bused on the radial basis function (RBF) network model and immune algorithm (IA) for continuous-time nonlinear systems. The nonlinear term of the system is represented by the RBF network. The IA is effectively introduced in order to track the time-varying system parameters and nonlinear term. The objective function for the identification is regarded as the antigen. The candidates of the estimated model are coded into binary bit strings as the antibodies and searched by the IA. Simulation results are shown to illustrate the proposed method.