高维多目标问题的排序新方法

doi:10.3969/j.issn.1001-2400.2014.06.015

J4 ›› 2014, Vol. 41 ›› Issue (6): 89-94.doi: 10.3969/j.issn.1001-2400.2014.06.015

高维多目标问题的排序新方法

代才¹;王宇平¹;何晓光²

(1. 西安电子科技大学计算机学院，陕西西安 710071；
2. 武警工程大学军事基础教育学院，陕西西安 710038)

收稿日期:2013-08-08 出版日期:2014-12-20 发布日期:2015-01-19
通讯作者: 代才
作者简介:代才(1984-), 男, 西安电子科技大学博士研究生, E-mail: daicai8403@126.com．
基金资助:
国家自然科学基金资助项目(61272119 和61203372)

New ranking method for many-objective problems

DAI Cai¹;WANG Yuping¹;HE Xiaoguang²

(1. School of Computer Science and Technology, Xidian Univ., Xi'an 710071, China;
2. Military Basic Institute of Education, Engineering University of CAPF, Xi'an 710038, China)

Received:2013-08-08 Online:2014-12-20 Published:2015-01-19
Contact: DAI Cai

摘要/Abstract

摘要：

针对高维多目标优化问题，提出了一种新的排序方法．它通过产生近似最优目标向量来增加种群的规模，从而达到对真实个体的有效排序．首先构造一个理想的帕累托前沿面，然后将这个理想的帕累托前沿面分成若干个网格，使每个个体都对应惟一的一个网格，通过这个网格上的节点来判断这个体是不是非支配解．数值实验表明，即使对于50维目标的问题，收敛性度量值也小于1．此外，与当前的两种最具代表性的松弛的帕累托占优方法比较，该方法能同时保持解的多样性和收敛性．

关键词: 高维多目标优化, 进化算法, 排序, 帕累托最优

Abstract:

A new ranking method is proposed to solve many-objective optimization problems. It can be used to generates many approximate optimal objective vectors to increase the size of population, so that the individuals can be efficiently sorted by using non-dominance sorting. An ideal Pareto Front is first constructed, and then the ideal Pareto Front is divided into a number of grids. For each individual, it is uniquely assigned to a grid, and then some nodes of the grid are used to determine whether the individual is a non-dominance solution. Experimental results show that, even for 50-objective optimization problems, the convergence measurements of solutions obtained by the ranking method are smaller than 1. Meanwhile, compared with two state-of-the-art relaxed forms of Pareto dominance, the experimental results show that this ranking method can simultaneously maintain the diversity of solutions and have good convergence.

Key words: many-objective optimization, evolutionary algorithms, ranking, Pareto optimisation

中图分类号:

代才;王宇平;何晓光. 高维多目标问题的排序新方法[J]. J4, 2014, 41(6): 89-94.

DAI Cai;WANG Yuping;HE Xiaoguang. New ranking method for many-objective problems[J]. J4, 2014, 41(6): 89-94.

参考文献

［1］ Kokshenev I, Braga A P. An Efficient Multi-objective Learning Algorithm for RBF Neural Network ［J］. Neurocomputing, 2010, 73(16-18): 2799-2808.
［2］ Depolli M, Trobec R, Filipc B. Asynchronous Master-Slave Parallelization of Differential Evolution for Multi-objective Optimization ［J］. Evolutionary Computation, 2013, 21(2): 261-291.
［3］ Liao H L, Wu Q H. Multi-objective Optimization by Learning Automata ［J］. Journal of Global Optimization, 2013, 55(2): 459-487.
［4］ Ishibuchi H, Tsukamoto N, Nojima Y. Evolutionary Many-objective Optimization: a Short Review ［C］//IEEE Congress on Evolutionary Computation. Piscataway: IEEE, 2008: 2419-2426.
［5］ Bader J, Zitzler E. HypE: an Algorithm for Fast Hypervolume-based Many-objective Optimization ［J］. Evolutionary Computation, 2011, 19(1): 45-76.
［6］ Friedrich T, Horoba C, Neumann, F. Multiplicative Approximations and the Hypervolume Indicator ［C］//Proceedings of the 11th Annual Genetic and Evolutionary Computation Conference. New York: ACM, 2009: 571-578.
［7］ Zitzler E, Künzli S. Indicator-based Selection in Multiobjective Search ［C］//Lecture Notes in Computer Science: 3242. Berlin: Springer, 2004: 832-842.
［8］ Ray T, Asafuddoula M D, Isaacs A. A Steady State Decomposition Based Quantum Algorithm for Many Objective Optimization ［C］//IEEE Congress on Evolutionary Computation. Piscataway: IEEE, 2013: 2817-2824.
［9］ Kokolo I, Hajime K, Shigenobu K. Failure of Pareto-based MOEAs: Does Nondominated Really Mean Near to Optimal? ［C］//Proceedings of Congress on Evolutionary Computation. Piscataway: IEEE, 2001: 957-962.
［10］马晶晶, 杨咚咚, 焦李成. 免疫非支配自适应粒子群多目标优化［J］. 西安电子科技大学学报, 2010, 37(5): 846-851.
Ma Jingjing, Yang Dongdong, Jiao Licheng. Immune Nondominated Adaptive Particle Swarm Multi-objective Optimization ［J］. Journal of Xidian University, 2010, 37(5): 846-851.
［11］ Sato H, Aguirre H E, Tanaka K. Controlling Dominance Area of Solutions and Its Impact on the Performance of mOEAs ［C］//Lecture Notes in Computer Science: 4003. Heidelberg: Springer, 2007: 5-20.
［12］ Deb K. Multi-objective Optimization Using Evolutionary Algorithms ［M］. New Jersey: John Wiley ＆ Sons, 2001.
［13］ Deb K, Agrawal S, Pratap A, et al. A Fast and Elitist Multiobjective Genetic Algorithm: NSGA-Ⅱ ［J］. IEEE Transactions on Evolutionary Computation, 2002, 6(2):182-197.
［14］ Coello C A, Cortés N C. Solving Multiobjective Optimization Problems Using an Artificial Immune System ［J］. Genetic Programming and Evolvable Machines, 2005, 6(2): 163-190.
［15］ Deb K, Thiele L, Laumanns M, et al. Scalable Multi-objective Optimization Test Problems ［C］//Proceedings of Congress on Evolutionary Computation. Piscataway: IEEE, 2002: 825-830.
［16］ Tan Yanyan, Jiao Yongchang, Li Hong, et al. MOEA/D+uniform Design: a New Version of MOEA/D for Optimization Problems with Manyobjectives ［J］. Computers ＆ Operations Research, 2013, 40(6):1648-1660.

[1]	谭雯,甘新标,白皓,肖调杰,陈旭光,雷书梦,刘杰. 面向超级计算机系统的大规模图遍历优化[J]. 西安电子科技大学学报, 2021, 48(6): 84-95.
[2]	杨五四,陈莉,王毅,张茂省. 一种适应度排序的高维多目标粒子群优化算法[J]. 西安电子科技大学学报, 2021, 48(3): 78-84.
[3]	周杰,徐梦颖,王娇娇,卢毅. 认知无线传感器网络频谱分配的一种改进方法[J]. 西安电子科技大学学报, 2020, 47(3): 80-85.
[4]	刘明骞,孟燕,张卫东. 雷达辐射源信号识别的效能综合评估方法[J]. 西安电子科技大学学报, 2019, 46(6): 1-8.
[5]	杜振鑫,刘广钟,赵学华. 集成学习人工蜂群算法[J]. 西安电子科技大学学报, 2019, 46(2): 124-131.
[6]	宋嘉奇, 陶海红. 一种稀布阵列天线的近场波束综合算法[J]. 西安电子科技大学学报, 2018, 45(6): 14-18.
[7]	李枫, 王紫阳, 王乐, 李蕊, 冯罡, 尹应增. 抗干扰阵列天线的小型化及波束形成技术分析[J]. 西安电子科技大学学报, 2018, 45(6): 51-56.
[8]	杨国平;刘三阳;张建科. 极小极大问题的生物地理学优化邻近点算法[J]. 西安电子科技大学学报, 2016, 43(5): 88-92+182.
[9]	刘娟妮;周诠. SMVQ索引分类编码无损信息隐藏算法[J]. 西安电子科技大学学报, 2016, 43(5): 128-132+146.
[10]	雷雨;焦李成;公茂果;李玲玲. 求解多目标考试时间表问题的NNIA改进算法[J]. 西安电子科技大学学报, 2016, 43(2): 157-161+192.
[11]	温鸿航;葛建华;许唐雯. 一种低复杂度OFDM峰均比降低方法[J]. J4, 2015, 42(5): 188-193.
[12]	董宁;王宇平. 求解约束优化问题的偏好多目标进化算法[J]. J4, 2014, 41(1): 98-104+188.
[13]	彭亚丽;刘芳. 一种组织进化的遮挡点恢复算法[J]. J4, 2013, 40(4): 137-141.
[14]	盖炳帅;王劲林;刘学. 一种业务关联的数据采集方法[J]. J4, 2013, 40(2): 54-59+158.
[15]	盖炳帅王劲林刘学. 一种业务关联的数据采集方法[J]. J4, 2013, 40(2): 54-59.

高维多目标问题的排序新方法

New ranking method for many-objective problems

PDF (PC)

赞

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 0