电子学报 ›› 2021, Vol. 49 ›› Issue (6): 1041-1049.DOI: 10.12263/DZXB.20201144

所属专题: 粒子群优化算法

• 学术论文 •    下一篇

基于直觉模糊Memetic框架的双粒子群混合优化算法

王毅1, 王侃琦1, 张茂省2, 李静1   

  1. 1. 西北大学信息科学与技术学院, 陕西西安 710127;
    2. 西安交通大学, 陕西西安 710049
  • 收稿日期:2020-10-18 修回日期:2021-01-06 出版日期:2021-06-25
    • 通讯作者:
    • 张茂省
    • 作者简介:
    • 王毅 男,1979年2月出生,上海人,博士(后).现为西北大学信息科学与技术学院副教授,主要从事智能信息处理、深度学习与群体智能优化;王侃琦 男,1994年出生,浙江杭州人.西北大学信息科学与技术学院硕士研究生,主要研究方向为群体智能优化算法;李静 女,1997年1月出生,山西晋中人.西北大学信息科学与技术学院硕士研究生,主要研究方向为多目标优化,深度学习.
    • 基金资助:
    • 国家重点研发计划 (No.2018YFC1504705); 国家重点自然科学基金 (No.61731015); 国家自然科学基金重大仪器专项 (No.42027806); 国家自然科学基金 (No.61402517); 国家重点实验室基金 (No.2016ADL-DW0302); 陕西省自然科学基金 (No.2018JM6029)

Hybrid Double Particle Swarm Optimization Algorithm Based on Intuitionistic Fuzzy Memetic Framework

WANG Yi1, WANG Kan-qi1, ZHANG Mao-sheng2, LI Jing1   

  1. 1. School of Information Science and Technology, Northwestern University, Xi'an, Shaanxi 710127, China;
    2. Xi'an Jiaotong University, Xi'an, Shaanxi 710049, China
  • Received:2020-10-18 Revised:2021-01-06 Online:2021-06-25 Published:2022-06-25
    • Corresponding author:
    • ZHANG Mao-sheng
    • Supported by:
    • National Key Research and Development Program of China (No.2018YFC1504705); National Natural Science Foundation of China (No.61731015); Major Instrument research projects of National Natural Science Foundation of China (No.42027806); National Natural Science Foundation of China (No.61402517); State Key Laboratory Foundation (No.2016ADL-DW0302); Natural Science Foundation of Shaanxi Province,  China (No.2018JM6029)

摘要: 为了平衡粒子群算法多样性与收敛速度,本文在Memetic框架下结合多属性决策,提出基于直觉模糊Memetic双种群混合优化算法.算法采用探索与开发分布式策略,在探索阶段,设计了社会强化算子和碰撞反弹算子提升种群多样性与勘探更多新区域;通过建立直觉模糊多属性决策对探索区域综合评估并生成可能存在的全局最优解区域,进而指导具有拉马克学习的开发种群进行局部精细搜索,实现不同策略下种群间的分布式协作与计算资源的合理分配.通过与其它5种新型进化算法在23个基准函数测试结果中体现出本算法具有更好的综合优化能力.

关键词: 粒子群, Memetic框架, 直觉模糊多属性决策, 分布式协作, 拉马克学习

Abstract: In order to mitigate the difficulty of balancing diversity and convergence in heuristic algorithm, this paper proposes an IF-memetic hybrid double particle swarm optimization (IFMHDPSO) based on intuitionistic fuzzy memetic framework and multi-attribute decision. There are two independent exploration and exploitation populations employing distributed strategies in which social reinforcement operator and collision rebound operator are proposed to improve diversity of algorithm and explore new areas in populations of exploration. Moreover, an intuitionistic fuzzy multi-attribute decision making is built up for comprehensively evaluating the solution space to get the potential global optimal solution area, which can guide the PSO (Particle Swarm Optimization) with Lamarckian mechanism to carry out the local search to achieve cooperation between populations under different strategies and reasonable allocation of computational resources. Compared with other 5 new evolutionary algorithms, IFMHDPSO is of better comprehensive optimization in 23 benchmark function test results.

Key words: particle swarm optimization, memetic framework, intuitionistic fuzzy multi-attribute decision making, distributed cooperation, Lamarckian learning

中图分类号: