Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (23): 38-40.DOI: 10.3778/j.issn.1002-8331.2010.23.010

• 研究、探讨 • Previous Articles     Next Articles

Self-adaptive particle swarm optimizer for multi-objective optimization problems

WEN Ying,LIAO Wei-zhi,BI Ying-zhou   

  1. School of Computer and Information Engineering,Guangxi Teachers Education University,Nanning 530023,China
  • Received:2009-05-19 Revised:2009-07-08 Online:2010-08-11 Published:2010-08-11
  • Contact: WEN Ying

求解多目标优化问题的自适应粒子群算法

文 瑛,廖伟志,闭应洲   

  1. 广西师范学院 计算机与信息工程学院,南宁 530023
  • 通讯作者: 文 瑛

Abstract: In this paper,a new multi-objective particle warm optimizer based on self-adaptive inertia weight is proposed.The particles are given different inertia weight based on fitness allocated by a new approach in order to control global exploring and local exploiting.The non-dominated solutions are archived external and crowding distance is used to maintain diversity.Meanwhile the strategy of elitist individual migration and turbulence enhance the convergence speed.The experimental results indicate that the proposed approach is competitive,being able to approximate the Pareto front efficiently.

Key words: multi-objective optimization, Particle Swarm Optimization(PSO), inertia weight, Pareto optimal

摘要: 提出了一种基于自适应惯性权重的多目标粒子群优化算法AWMOPSO,采用新的适应值分配机制,在搜索过程中根据粒子的适应值对粒子进行分类,动态调整粒子的惯性权重以控制粒子的开发和探索能力。用外部精英集保存非支配解,并通过拥挤距离维持解的多样性。引入精英迁移和局部扰动策略,提高收敛的速度和精度。典型的测试函数的计算结果表明了算法能够快速逼近Pareto最优前沿,是求解多目标优化问题的有效方法。

关键词: 多目标优化, 粒子群, 惯性权重, Pareto最优

CLC Number: