A Review and Empirical Analysis of Particle Swarm optimization Algorithms for Dynamic Multi-Modal optimization
Simon Dennis, Andries Petrus Engelbrecht · 2020
A number of particle swarm optimization (PSO) variations have been developed to find multiple solutions to multimodal optimization problems. These algorithms have been extensively evaluated in the literature. When dynamic optimization problems are considered, only a few PSO algorithms exist that have the ability to find and track multiple optima in dynamically changing search landscapes. These algorithms have not yet been rigorously evaluated on an extensive set of dynamic optimization problems. This paper presents a review of existing dynamic multimodal PSO algorithms and conducts an empirical analysis of these algorithms on a set of dynamic optimization problems of varying dynamics. The best performing dynamic multi-modal PSO algorithms, with respect to different performance measures, are identified as an outcome of a formal statistical analysis.