Performance Comparison of Partical Swarm Optimization Variant Models

Abstract

In this work, an extensively comparative study is conducted to demonstrate the performance of Particle Swarm Optimization (PSO) variants based on five well-known benchmark functions in the area. According to the PSO's cognitive and social factors' contribution, we categorize the PSO algorithm into five variants. Different from other research work, which included only four PSO models, we propose an extra PSO variant called selfless Full-Model. Therefore, the five PSO variants, which named PSO Full-Model, PSO Cognitive-Only Model, PSO Social-Only Model, PSO Selfless Model and PSO Selfless Full-model, respectively, are applied to solve the benchmark functions. Their performances are compared based on the success rate, average function evaluations and the best fitness.