电子学报 ›› 2018, Vol. 46 ›› Issue (2): 315-324.DOI: 10.3969/j.issn.0372-2112.2018.02.009

• 学术论文 • 上一篇    下一篇

一种基于多样性信息和收敛度的多目标粒子群优化算法

韩红桂1,2, 卢薇1,2, 乔俊飞1,2   

  1. 1. 北京工业大学信息学部, 北京 100124;
    2. 计算智能与智能系统北京市重点实验室, 北京 100124
  • 收稿日期:2016-08-07 修回日期:2016-11-07 出版日期:2018-02-25
    • 作者简介:
    • 韩红桂,男,1983年8月出生,江苏泰州人.北京工业大学教授、博士生导师.主要研究领域为城市污水处理过程智能优化控制、计算智能与智能系统等.E-mail:rechardhan@bjut.edu.cn;卢薇,女,1993年8月出生,山东日照人.2015年本科毕业于菏泽学院,并于2015年进入北京工业大学信息学部,现为硕士研究生.主要研究领域为污水处理过程智能优化和神经网络.E-mail:luweiwei@emails.bjut.edu.cn
    • 基金资助:
    • 国家自然科学基金 (No.61622301,No.61533002); 中国博士后科学基金 (No.2014M550017); 教育部博士点基金 (No.20131103110016); 北京市教育委员会项目 (No.km201410005001,No.KZ201410005002)

A Multiobjective Particle Swarm Optimization Algorithm Based on the Diversity Information and Convergence Degree

HAN Hong-gui1,2, LU Wei1,2, QIAO Jun-fei1,2   

  1. 1. Department of Information, Beijing University of Technology, Beijing 100124, China;
    2. Beijing Key Laboratory of Computational Intelligence and Intelligent System, Beijing 100124, China
  • Received:2016-08-07 Revised:2016-11-07 Online:2018-02-25 Published:2018-02-25
    • Supported by:
    • National Natural Science Foundation of China (No.61622301, No.61533002); China Postdoctoral Science Foundation (No.2014M550017); Ph.D. Programs Foundation of Ministry of Education of China (No.20131103110016); Project of Beijing Municipal Education Commission (No.km201410005001, No.KZ201410005002)

摘要: 为了提高多目标粒子群算法优化解的多样性和收敛性,提出了一种基于多样性信息和收敛度的多目标粒子群优化算法(Multiobjective Particle Swarm Optimization based on the Diversity Information and Convergence Degree,dicdMOPSO).首先,利用非支配解多样性信息评估知识库中最优解的分布状态,设计出一种全局最优解选择机制,平衡了种群的进化过程,提高了非支配解的多样性和收敛性;其次,基于种群多样性信息设计出一种飞行参数调整机制,增强了粒子的全局探索能力和局部开发能力,获得了多样性和收敛性较好的种群.最后,将dicdMOPSO应用于标准测试函数测试,实验结果表明,dicdMOPSO与其他多目标算法相比不仅获得了多样性较高的可行解,而且能够较快的收敛到Pareto前沿.

关键词: 智能优化算法, 多目标粒子群优化, 种群多样性信息, 非支配解多样性信息, 收敛度

Abstract: To improve the diversity and convergence of optimal solutions in multiobjective particle swarm optimization (MOPSO) algorithm, a multiobjective particle swarm optimization algorithm, based on the diversity information and convergence degree, named dicdMOPSO, is developed in this paper. Firstly, a global optimal solution selection mechanism, based on the distribution of optimal solutions in the knowledge base with the diversity information of non-dominated solutions,is introduced to balance the evolutionary process of population to improve the diversity and convergence of non-dominated solutions. Then, to enhance global exploration and local exploitation abilities of particles, a flight parameter adjustment mechanism is proposed to obtain the particles with better diversity and convergence by using the population diversity information. Finally, the experiment results demonstrate that, compared with other multiobjective algorithms, this proposed dicdMOPSO algorithm can not only obtain the optimal solutions with better diversity, but also be faster to catch the Pareto front.

Key words: intelligent optimization algorithm, multiobjective particle swarm optimization, population diversity information, diversity information of non-dominated solutions, convergence degree

中图分类号: