A new design method using opposition-based BAT algorithm for IIR system identification problem

Abstract

BAT algorithm (BA) is a meta-heuristic algorithm, based on the echolocation behaviour of bats. In this paper, optimal set of filter coefficients is searched by the modified optimisation methodology called opposition-based BAT algorithm (OBA) for infinite impulse response (IIR) system identification problem. Opposition based numbering concept is embedded into the primary foundation of BA metaphorically to enhance the convergence speed and performance for finding better near-global optimal solution. Detailed and balanced search in multidimensional problem space is accomplished with judiciously chosen control parameters of OBA technique. When tested against standard benchmark examples, for same and reduced order models, the simulation results establish the OBA as a more competent candidate to other evolutionary algorithms as real coded genetic algorithm (RGA), differential evolution (DE) and particle swarm optimisation (PSO) in terms of accuracy and convergence speed.