Improved identification of Hammerstein plants using new CPSO and IPSO algorithms

Abstract

Identification of Hammerstein plants finds extensive applications in stability analysis and control design. For identification of such complex plants, the recent trend of research is to employ nonlinear network and to train their weights by evolutionary computing tools. In recent 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 by combining the principles of AIS and PSO, we propose two new but simple hybrid algorithms called Clonal PSO (CPSO) and Immunized PSO (IPSO) which involve less complexity and offers better identification performance. Identification of few benchmark Hammerstein models is carried out through simulation study and the results obtained are compared with those obtained by standard PSO, Clonal and GA based methods. Various simulation results demonstrate that IPSO algorithm offers best identification performance compared to the other algorithms. Out of the two algorithms proposed, the CPSO is computationally simpler but offers identification performance nearly similar to its PSO counterpart.