Enhanced PSO Based on Multi-Agent System

Traditional particle swarm optimization (PSO) algorithm is combined with multi-agent system (MAS), so the particles are upgraded to intelligent agents, which are more autonomous and smart. Secondly, Evolutionary Programming (EP) is integrated into agents to improve the search capability of standard PSO particles by altering their intrinsic tendency of moving to global best position. Thirdly, an adaptive mechanism is proposed to lessen Vmax´s impact over algorithm performance, so that the algorithm becomes more feasible. Finally, the enhanced algorithm is applied in the optimization problem of complex functions of high dimension and satisfied results are achieved.