Akademik Veri Yönetim Sistemi
CIRCUITS SYSTEMS AND SIGNAL PROCESSING, cilt.44, sa.6, ss.4059-4081, 2025 (SCI-Expanded, Scopus)
The problem of infinite impulse response (IIR) system identification is a key and demanding issue in signal processing and control systems, which requires appropriate system dynamics describing a model that also ensures the system's stability and computational efficiency. In this study, a modified atom search optimizer (mASO) is proposed, combining an adaptive gbest-guided mechanism to overcome weaknesses in the standard atom search optimizer. Through this proposed mASO, time wastes that are related to poor local optimum and loss of control over exploration and exploitation processes are tamed, meaning that this idea is suited for difficult instances of OCO. The algorithm was tested on second to fifth-order systems and employed the same-order models as well as the reduced-order IIR models. These analyses include mean square error (MSE) estimates, pole-zero diagram checks, and statistical tests for significance for a variety of factors. Several optimization methods were employed in order to compare the performance of mASO, including ASO, moth flame optimizations, particle swarm optimizations, inclined plane system optimization, gravitational search algorithm, genetic algorithm, selfish herd optimizations, and many other contemporary ones. In all the systems mentioned, mASO was found to have performed well in comparison, noting that average MSE values of the second order same order model were as high as 7.8072E-34 and 3.6286E-03 for the fourth order reduced order model. Details of pole-zero diagrams confirmed stability and dynamic accuracy, while the statistical results indicated that mASO is consistent and robust under varying systems. The findings point out the mASO's capacity to be used in complex IIR system identification tasks with high efficiency. Its proven success against other algorithms makes the mASO a credible and formidable technique for several engineering works, including the design of robust filters and the modeling of dynamic systems.