Study On Multi-objective Particle Swarm Optimization Algorithm And Its Application

Multi-objective optimization problems are widespread in many areas of real life, andmany algorithms have been proposed for solving them. With the constant development ofoptimization techniques, multi-objective optimization algorithms have experienced a growingprocess from traditional multi-objective optimization algorithms to evolutionary algorithms.Due to its simplicity of implementation compared to other evolutionary algorithms, PSO hasits own advantages in dealing with multi-objective optimization problem, so multi-objectiveparticle swarm optimization has been gaining increasing attention in recent years. Based onthe theory of optimization, particle swarm optimization and multi-objective optimizationproblem are studied as the main subjects of this paper, and the multi-objective particle swarmoptimization algorithm and its application are discussed. This paper is organized as follows:In chapter one, we elaborate on the background and significance of the problems we willstudy, introduce the mathematical models of single-object optimization problem andmulti-objective problem respectively and some basic concepts pertaining to multi-objectiveproblem; then we summarize the traditional multi-objective optimization algorithms andmulti-objective evolution algorithms respectively, and describe some representativealgorithms. The main tasks and innovations of this paper are finally introduced.In chapter two, basic particle swarm optimization algorithm is introduced, the basicprocedures of particle swarm optimization algorithm are given, and several improved particleswarm optimization algorithm are expounded. The main focus of this chapter is that itproposes an improved particle swarm optimization algorithm, i.e., the Self-adaptive velocityPSO algorithm based on ring neighborhood topology, and evaluates the performance of thisalgorithm on some typical benchmark test functions.In chapter three, we introduce the basic theory of multi-objective optimization problemfirstly, including external set and the selection of global leader, and then we describe the basicprocess of multi-objective particle swarm optimization algorithm. The performancemeasurements of multi-objective particle swarm optimization algorithm are given finally.In chapter four, the concepts of the constrained multi-objective particle swarmoptimization algorithm are introduced, and a new constrained multi-objective particle swarmoptimization algorithm based on the improved algorithm proposed in chapter two. Later, wegive the implementation steps of this algorithm and validate it on some basic test functions.In chapter five, we apply the proposed multi-objective particle swarm optimizationalgorithm to the design of ship lines. We briefly introduce the concepts and applicationbackgrounds of ship lines design, give some objective functions in the design of large ship,then establish a reasonable mathematical model and solve it using the constrainedmulti-objective particle swarm optimization algorithm proposed in chapter four.In the last chapter, the whole research work of this paper is summarized, and somedirections of further research are pointed out.