Time Varying System Identification with Immune Based Evolutionary Computation

In this paper, an application of the immune mechanism based evolutionary algorithm to system identification of time varying systems and nonlinear systems is considered. The immune mechanism based evolutionary algorithm is one of evolutionary search or optimization algorithm, inspired by natural immune system. The immunological memory is used for the model preserving in the memory based modeling technique, the immunological response is applied to explore model candidate from population pool and memory cell. Antibodies and antigens are referred as the model candidates and the obtained sets of data, respectively. Then, the density of the antibodies is introduced based on Kullback information criterion and the model candidates are updated step by step with evolutionary computation. Some numerical studies are carried out to illustrate the applicability of the proposed algorithm to switching linear system and nonlinear system identification