Font Size: a A A

Improvement Of Multi-objective Particle Swarm Optimization Algorithm

Posted on:2020-09-05Degree:MasterType:Thesis
Country:ChinaCandidate:L S XuFull Text:PDF
GTID:2428330590962869Subject:Computational Mathematics
Abstract/Summary:PDF Full Text Request
Many engineering and practical problems can be summed up as multi-objective problems.We know that particle swarm optimization(PSO)has the characteristics of strong global search performance,fast convergence speed and concise concept,so it is widely used to solve the problem of large search space such as multi-target.For solving multi-objective problems,we now face two problems.The first is the degree of convergence of the optimal solution set to the real Pareto optimal solution,and the other is the uniformity of the distribution of the optimal solution set at the Pareto front.Therefore,this dissertation also improves the particle swarm optimization algorithm in two aspects: on the one hand,it proposes multi-objective optimization of chaotic particle swarm optimization algorithm: using logistic mapping in chaotic sequence to generate solutions after particle swarm updating;introducing crossover operator of normal distribution to improve the diversity of the population in the middle and later stages of the algorithm;and using simplified grid reduction and gene exchange.And other strategies to improve the performance of the algorithm.Finally,the experimental results are obtained by comparing with MOPSO,NSGA-II and MOEA/D algorithms.On the other hand,aiming at the strong global search performance and poor local search performance of particle swarm optimization algorithm and the strong local search performance of differential evolution algorithm,a multi-objective particle swarm cooperative differential evolution algorithm is proposed.Its core idea is to generate scale factor by simulated annealing algorithm according to the number of iterations,and dynamically adjust the population according to the scale factor.In order to reduce the computational complexity,the improved particle swarm optimization(PSO)algorithm and differential evolution(DE)algorithm are used to update the ratio,and the non-dominant solution set is reduced according to the crowding distance sorting method.Through test cases,it is compared with MOPSO,NSGA-II and some literature algorithms,and the performance of the algorithm is discussed.Finally,the dissertation summarizes and prospects the main work and conclusions of this dissertation,and gives the shortcomings of this chapter and the direction of improvement.
Keywords/Search Tags:Multi-objective optimization, Particle swarm algorithm, Differential evolution algorithm, Simulated annealing algorithm, Mesh reduction
PDF Full Text Request
Related items