A Hybrid Genetic and Particle Swarm Algorithm for Service Composition

摘要

Web Service Composition (WSC) has become a hotspot in recent research. Current solutions focus on ontology information representation and ontology based web service matching, which lacks flexibility. From simulation of human cognision, this paper proposed a hybrid Genetic Particle Swarm Algorithm (GPSA) to solve the problem of WSC, which is a Multi-Objective Problem (MOP). Genetic Algorithm (GA) is used to search throughout the problem space, and Particle Swarm Optimization (PSO) is used to enhance local search ability. PSO can reduce the calculation cost by trimming useless braches. Feedback information is used to decide how to balance GA and PSO, which means how to balance global and local optimization. Experiments show that GPSA can solve WSC Problem (WSCP) and balance between global and local optimization.