Empirical analysis and improvement of the PSO-sono optimization algorithm

¹ College of Computing and Systems, Abdullah Al Salem University, Khaldiya, Kuwait
² School of Information Engineering, Nanchang Institute of Technology, Nanchang 330099, P.R. China
³ Computer Science Department, Gulf University for Science & Technology, West Mishref, Kuwait

^* Corresponding author: omran.m@ieee.org

Received: 4 March 2024
Accepted: 16 March 2025

Abstract

PSO-sono is a recent and promising variant of the Particle Swarm Optimization (PSO) algorithm. It outperforms other popular PSO variants on many benchmark test sets. In this paper, we investigate the performance of PSO-sono on more problems (including 21 real-world optimization problems). Moreover, we propose a new, more powerful yet simpler and more efficient variant of PSO-sono, called IPSO-sono. The proposed approach uses ring topology, non-linear ratio reduction and opposition-based learning to improve the performance of PSO-sono. The proposed approach is compared with other state-of-the-art metaheuristic algorithms on 12 IEEE CEC 2022 and 21 real-world problem defined in the IEEE CEC 2011. The results show that IPSO-sono outperforms PSO-sono on most problems and performs well compared to other state-of-the-art approaches.