Assessing Performance of Morphology Particle Swarm Optimization (Morph-PSO) on Standard Benchmark Functions

The search boundary is a critical aspect of optimization problems. Understanding the limits of the search boundary can help prevent solutions from exiting the search area and assist optimization algorithms in converging more rapidly. Typi-cally, optimization algorithms require manual setting of boundary limits before execution. Consequently, many researchers are exploring adaptive approaches to address this challenge. This paper introduces a self-limiting behavior for Particle Swarm Optimization, inspired by Hooke's Law and names this algorithm Morphological Particle Swarm Optimization (Morph-PSO). Additionally, the study presents an extension of Morph-PSO, wherein a constriction factor, K, is integrated, resulting in the Constricted Morphological Particle Swarm Optimization (CMorph-PSO). These two novel methods were tested on thirteen selected benchmark functions to assess their performance. Morph-PSO appeared as the top-performing algorithm, achieving top performance in seven out of the selected benchmarks, compared to a range of alter-native optimization methods. CMorph-PSO did not perform as well, primarily due to the small step size affected by the double constriction factor (morphology and constriction). However, both proposed methods demonstrated high performance in convergence time.