A combined double chaotic maps and subswarm based enhanced particle swarm optimization for global optimization and engineering design problems

In this study, newly introduced a double chaotic maps based enhanced Particle Swarm Optimization (DCMEPSO) algorithm with two subswarm is presented to obtain a global optimum solution. In this context, Squared Sine Logistic (SSL) map for one of subswarm and Modular Integrated Logistic Exponential (MILE) map for other subswarm are presented to enrich population diversity and enhance global search capabilities. In addition, our method introduces different position updating mechanisms to facilitate the escape from possible local optima and speed search processes in achieving global optimum. The performance of the presented optimization algorithm has been investigated under eight well-known benchmark functions, 3 of which are unimodal and 5 of which are multimodal. We consider four optimization algorithms such as Particle Swarm Optimization (PSO), Whale Optimization Algorithm (WOA), Grey Wolf Optimization (GWO), and Salp Swarm Algorithm (SSA), as competitive algorithms to assess the performance of DCMEPSO with experimental studies such as well-known functions of different sizes, and engineering design problems. The superiority of the proposed algorithm from other compared algorithms is demonstrated with Wilcoxon rank sum test to possess the definitive assessment. The proposed algorithm can easily tackle with high-dimensional complex problems. From extensive test results, mean rank analysis over 8 benchmark functions shows that the proposed method achieves 60.45% and 50.0% mean-rank-based improvements in the 30D and 50D cases, respectively, compared to the closest competitor. The Matlab code of DCMEPSO is available at: https://github.com/fevzeddinulker/DCMEPSO/ .