Multi-objective Particle Swarm Optimization Method Based on Fitness Function and Sequence Approximate Model

Heuristic search methods usually require a large amount of evolutionary iterative calculation, which has become a bottleneck for applying them to practical engineering problems. In order to reduce the number of analysis of heuristic search methods, a pareto Multi-objective Particle Swarm Optimization(MOPSO) method is presented. In this approach, pareto fitness function is used to select global extremum particles. And the solution accuracy and efficiency are balanced by adopting sequence approximate model. Research shows that the method can ensure the accuracy of calculation, at the same time help to reduce the number of accurate analysis.

I. Introduction

As one of heuristic search methods, Particle Swarm Optimization(PSO) [1] is a population based technique, which have made it a natural candidate to be extended for MultiObjective Optimization(MOO). The key technologie of MOPSO lies in the selection strategy of global and personal best particle position, which directly affectes the convergence to the true Pareto front as well as producing a well distributed Pareto front are. Therefore the existing MOPSOs are mainly centered on studying the abovementioned key technologie. According to different ways of choice best particle position, existing MOPSO methods are mainly divided into the following categories: