SYSTEM IDENTIFICATION USING HAMMERSTEIN MODEL OPTIMIZED WITH ARTIFICIAL BEE COLONY ALGORITHM


Özer Ş.

ÖHÜ Müh. Bilim. Derg., vol.7, no.1, pp.1-16, 2018 (Peer-Reviewed Journal)

  • Publication Type: Article / Article
  • Volume: 7 Issue: 1
  • Publication Date: 2018
  • Journal Name: ÖHÜ Müh. Bilim. Derg.
  • Journal Indexes: TR DİZİN (ULAKBİM)
  • Page Numbers: pp.1-16

Abstract

 Hammerstein model is formed by cascade of linear and nonlinear parts. In literature, memoryless polynomial nonlinear (MPN) model for nonlinear part and finite impulse response (FIR) model or infinite impulse response (IIR) model for linear part are mostly preferred for Hammerstein models. This paper different from the studies in literature, focuses on the success of Hammerstein block model that Second Order Volterra (SOV) is preferred instead of MPN as nonlinear part. In this context, a new Hammerstein model is presented which is obtained by cascade form of a nonlinear SOV and a linear FIR model. In simulations, different types of system are identified by proposed Hammerstein model which is optimized with ABC (artificial bee colony) algorithm. The simulation results reveal effectiveness and robustness of the proposed model with ABC algorithm.