Adaptive infinite impulse response system identification using an enhanced golden jackal optimization

Abstract

Abstract The significant error of the adaptive infinite impulse response (IIR) system identification often involves nonlinearity and indifferentiability, the practical and reliable swarm intelligence optimization techniques are required to calculate and establish the ideal parameters of the IIR filter. In this research, an enhanced golden jackal optimization (EGJO) based entirely on the elite opposition-based learning technique and the simplex technique can be adopted to address this issue. The intention is to minimize the error fitness value and attain the appropriate control parameters. The golden jackal optimization (GJO), based on the cooperative attacking behavior of the golden jackals, simulates the searching for prey, stalking and enclosing prey, pouncing prey to efficaciously tackle the complicated optimization problem. The elite opposition-based learning technique has the characteristics of boosting population diversity, enhancing exploration ability, extending search range and avoiding search stagnation. The simplex technique has the characteristics of accelerating the search process, enhancing the exploitation ability, improving the computational precision and increasing the optimization depth. The EGJO can realize the balance between exploration and exploitation to arrive at the best possible outcome. To demonstrate the overall search ability, the EGJO is compared with those of the AOA, GTO, HHO, MDWA, RSO, WOA, TSA and GJO by gradually decreasing the error fitness value of the IIR filter. The experimental results clearly demonstrate that the optimization efficiency and recognition accuracy of EGJO are superior to those of other algorithms. The EGJO offers several benefits to obtaining a faster convergence rate, higher computation precision, better control parameters and better fitness value. In addition, the EGJO is very stable and resilient in tackling the IIR system identification problem.