Identification of a Box-Jenkins structured two stage cascaded model using Simplex Particle Swarm Optimization algorithm

Abstract

This paper delivers an efficient and accurate approach for identification of a Box-Jenkins (BJ) structure based Wiener model with Simplex Particle Swarm Optimization (SPSO) algorithm. The accuracy and the precision of the identification scheme have been justified with the reported bias and variance information, respectively, of the estimated parameters. The output mean square error (MSE) has been considered as the fitness function to be optimized for the SPSO algorithm. The accuracy and the consistency of the identification of the Hammerstein system have been justified with the corresponding statistical information of the MSE. Accurate identification of the parameters associated with the linear dynamic sub-system ensures the stability of the overall closed loop system.