Modified Firefly Optimization for IIR System Identification

Abstract

Because of the nonlinear and recursive nature of the physical systems, system identification is a challenging and complex optimization problem. Infinite impulse response (IIR) and nonlinear adaptive systems are widely used in modeling real-world systems. IIR models due to their reduced number of parameters and better performance are preferred over finite impulse response (FIR) systems. In the past few decades, meta-heuristic optimization algorithms have been an active area of research for solving complex optimization problems. In the present paper a modified version of a recently introduced population-based firefly algorithm (MFA) is used to develop the learning rule for identification of three benchmark IIR and nonlinear plants. MFA's performance is compared with standard firefly algorithm (FA), GA and three versions of PSO. The results demonstrate that MFA is superior in identifying dynamical systems.