计算机应用 ›› 2012, Vol. 32 ›› Issue (11): 2985-2988.DOI: 10.3724/SP.J.1087.2012.02985

• 先进计算 • 上一篇    下一篇

通过网格改进的基于指标的进化算法

肖宝秋,刘洋,戴光明   

  1. 中国地质大学(武汉) 计算机学院,武汉 430074
  • 收稿日期:2012-05-21 修回日期:2012-06-25 发布日期:2012-11-12 出版日期:2012-11-01
  • 通讯作者: 肖宝秋
  • 作者简介:肖宝秋(1988-),男,江西吉安人,硕士研究生,主要研究方向:多目标优化、科学计算可视化;刘洋(1987-),女,黑龙江齐齐哈尔人,硕士研究生,主要研究方向:多目标优化;戴光明(1964-),男,湖北武汉人,教授,博士生导师,主要研究方向:算法设计与分析、科学计算可视化。
  • 基金资助:
    国家自然科学基金资助项目;湖北省自然科学基金资助项目(2011CDB348)

Improved indicatorbased evolutionary algorithm based on grid

XIAO Bao-qiu,LIU Yang,DAI Guang-ming   

  1. School of Computer Science, China University of Geosciences, Wuhan Hubei 430074,China
  • Received:2012-05-21 Revised:2012-06-25 Online:2012-11-12 Published:2012-11-01
  • Contact: XIAO Bao-qiu

摘要: 设计一种高效的演化多目标优化算法,使其能获得一组同时具有优异的收敛性和多样性的解集是一项很困难的任务。为了能高效求解多目标优化问题,在基于指标的进化算法(IBEA)的基础上:1)引入基于目标空间网格的多样性保持策略,保证算法近似前沿具有优异的分布性;2)引入反向学习机制,同时评估当前解和当前解的反向解,期望能找到一组较优的解从而加快算法收敛。通过6个标准测试函数对改进算法进行测试,其结果表明改进算法可以有效逼近真实Pareto前沿并且分布均匀。

关键词: 多目标优化, 基于指标的进化算法, 网格, 反向学习

Abstract: Evolutionary MultiObjective Optimization (EMO) has become a very popular topic in the last few years. However, designing an efficient EMO algorithm for finding wellconverged and welldistributed approximate optimal set is a challenging task. In this paper, an improved IBEA algorithm was proposed to solve Multiobjective Optimization Problems (MOPs) efficiently. The proposed approach introduced a diversity promotion mechanism based on grid in objective space to ensure the approximate optimal set has good distribution. To make the algorithm converge faster, the new approach employed oppositionbased learning mechanism to evaluate the current solutions and their opposite solutions simultaneously in order to find a group of better solutions. The experiments on six benchmark problems show that the new approach is able to obtain a set of welldistributed solutions approximating the true Paretofront.

Key words: multiobjective optimization, IndicatorBased Evolutionary Algorithm (IBEA), grid, oppositionbased learning

中图分类号: