Enhancing IIR system identification: Harnessing the synergy of gazelle optimization and simulated annealing algorithms

Abstract

In the field of digital filter design and system identification, accurately modeling Infinite Impulse Response (IIR) systems is of utmost importance. This paper introduces a new adaptive algorithm that combines the gazelle optimization algorithm with simulated annealing to achieve superior optimization performance for digital IIR filters used in system identification. To evaluate the algorithm's effectiveness, extensive experiments were conducted on two fronts: the challenging CEC2020 benchmark functions and a fifth order IIR system identification problem. The proposed algorithm proves its effectiveness by achieving optimal solutions on the CEC2020 benchmark functions when compared to other well-known optimization algorithms such as arithmetic optimization, particle swarm optimization, differential evolution, sine cosine algorithm, gray wolf optimizer, biogeography-based optimization, salp swarm algorithm, and the original gazelle optimization algorithm. The algorithm's performance is measured by its ability to find the best solutions for these benchmark functions. For the IIR system identification problem, the study considers both the same order and reduced order cases. The system to be identified is represented by a fifth order plant. The algorithm's statistical analysis and convergence behavior are compared to the original gazelle optimization algorithm, artificial hummingbird algorithm, mountain gazelle optimization algorithm, and the widely used artificial bee colony algorithm. These comparisons demonstrate the proposed algorithm's superior capability and efficiency in accurately identifying the parameters of IIR systems.