Transitional and Parallel Approach of PSO and SGO for Solving Optimization Problems

Abstract

Optimization is finding minimization or maximization of a decision variable for a given problem. In most of the engineering problems, there is a requirement of optimizing some variables or other to obtain a desired objective. Several classical techniques are existing in optimization literature. However, when the optimization problem is complex, discrete, or not derivable there is a need to look beyond classical techniques. Swarm intelligence techniques are overwhelmingly popular in current days for targeting such kind of optimization problems. Particle Swarm Optimization (PSO) and Social Group Optimization (SGO) are belonging to this category. PSO being a popular and comparatively an older algorithm to SGO, the efficiency and efficacy of PSO for function optimization are well established. In this paper, an effort is made to explore an effective alternate model in hybridizing PSO and SGO. In the proposed model, a transitional concept is used. An alternate switching between PSO and SGO is carried out after a fixed iteration. An exhaustive simulation is done on several benchmark functions and a comparative analysis is presented at the end. The results reveal that the proposed approach is a better alternative to obtain effective results in comparatively less iterations than stand-alone models.