A new Hybrid PSO-SCA using Horse Optimization Algorithm's group behavior update

Abstract

This paper proposes a new Hybrid Algorithm between Particle Swarm Optimization (PSO) and Sine Cosine Algorithm (SCA) with Horse Optimization Algorithm (HOA) group assigning and updating methodology. The proposed algorithm is called Hybrid PSO-SCA with HOA group behavior update (HPSH) aims to solve the disadvantage of both PSO and SCA. HPSH start by assigning each particle a group with the same methodology as HOA and then classifies them based on the fitness of their current position. Each group of particles will share the same movement equation which is different between groups. Particles are periodically assigned to the new group as the iteration increase, assigning criteria is based on fitness value. Movement equations of HPSH are the combination of PSO and SCA, which depend on assigned group. HPSH is experimented in 24 benchmark functions, assigning majority of test functions with 100 dimensions. The experimental results indicate that HPSH has retained both PSO and SCA on almost every function, while some outperform PSO and SCA.