基于相关分析的多目标优化Pareto优劣性预测

doi:10.3969/j.issn.0372-2112.2017.02.027

电子学报 ›› 2017, Vol. 45 ›› Issue (2): 459-467.DOI: 10.3969/j.issn.0372-2112.2017.02.027

所属专题：多目标优化

基于相关分析的多目标优化Pareto优劣性预测

李文彬^1,2, 贺建军¹, 郭观七², 冯彩英², 潘理²

1. 中南大学信息科学与工程学院, 湖南长沙 410083;
2. 湖南理工学院信息与通信工程学院, 湖南岳阳 414006

收稿日期:2015-05-11 修回日期:2016-07-25 出版日期:2017-02-25
- 通讯作者:
- 郭观七
- 作者简介:
- 李文彬,男,1981年出生,湖南岳阳人,现为中南大学信息科学与工程学院博士研究生,主要研究方向为进化计算、多目标优化、神经网络.E-mail:wenbin_lii@163.com;贺建军,男,1965年出生,湖南宁乡人,中南大学信息科学与工程学院教授、博士生导师.主要研究方向为复杂工业过程建模与优化控制、自适应控制理论与应用.E-mail:jjhe@mail.csu.eud.cn
- 基金资助:
- 国家自然科学基金 (No.60975049，No.61174132）; 湖南省省教育厅科学研究重点项目 (No.15A079）; 湖南省高校科技创新团队支持计划资助

Prediction of Pareto Dominance Based on Correlation Analysis

LI Wen-bin^1,2, HE Jian-jun¹, GUO Guan-qi², FENG Cai-ying², PAN Li²

1. College of Information Science and Engineering, Central South University, Changsha, Hunan 410083, China;
2. College of Information and Communication Engineering, Hunan Institute of Science and Technology, Yueyang, Hunan 414006, China

Received:2015-05-11 Revised:2016-07-25 Online:2017-02-25 Published:2017-02-25
- Supported by:
- National Natural Science Foundation of China (No.60975049, No.61174132); Science Research Key Program of Education Department of Hunan Province (No.15A079); Funded by Science and Technology Innovation Team Program of Hunan University

摘要/Abstract

摘要：

昂贵多目标进化算法中，目标向量评估所需计算时间或实验成本高昂，大量昂贵评估必然导致成本灾难.本文根据多目标优化Pareto优劣性取决于各目标分量的序关系这一关键性质，提出一种序拟合方法进行Pareto优劣性预测.在分析样本数据决策空间与目标空间序相关性的基础上，通过线性相关的假设条件，建立低成本的序关系预测方程，并用预测的序关系确定Pareto优劣性.然后对典型多目标优化问题进行Pareto优劣性预测对比实验，结果表明所提方法显著提高了Pareto优劣性的预测精度.最后，将该预测方法集成到NSGA-II算法中，可以避免进化过程中的模型重构，有效减少昂贵目标向量的评估次数.

关键词: 相关分析, 序关系预测, 多目标优化, Pareto优劣性

Abstract:

In expensive multi-objective evolutionary algorithms,the evaluation of a large number of objective vectors spend a lot of time or experimental cost and lead to the cost of disaster.According to the fact that Pareto dominance relationships among candidate solutions are depended on the rank relationships of objective components,this paper proposes a predict method of rank equivalent to determine Pareto dominance.A decision vector and object vector rank matrix is established,and rank correlation analysis is used to calculate the correlation coefficient matrix R.Under the assumption of linear correlation,a prediction equation is established to predict rank relationships.Testing results on typical multi-objective optimization problems show that the proposed method only requires establishing a linear prediction model,which can remarkably improve the prediction accuracy and reduce the calculation of original expensive target function.Finally,the prediction method is integrated into the NSGA-II,it can avoid reconstruction the model in the process of evolution,then effectively decrease the number of evaluation for expensive objective vectors.

Key words: correlation analysis, rank relation prediction, multi-objective optimization, Pareto dominance

中图分类号:

TP301

李文彬, 贺建军, 郭观七, 等. 基于相关分析的多目标优化Pareto优劣性预测[J]. 电子学报, 2017, 45(2): 459-467.

LI Wen-bin, HE Jian-jun, GUO Guan-qi, et al. Prediction of Pareto Dominance Based on Correlation Analysis[J]. Acta Electronica Sinica, 2017, 45(2): 459-467.

参考文献

[1] 詹炜.求解高维多目标优化问题的流形学习算法研究[D].武汉,中国地质大学,2013.
[2] Zhou A,Qu B Y,Li H,et al.Multiobjective evolutionary algorithms:A survey of the state of the art[J].Swarm & Evolutionary Computation,2011,1(1):32-49.
[3] Seah C W,Ong Y S,Tsang I W,et al.Pareto rank learning in multi-objective evolutionary algorithms[A].2012 IEEE Congress on Evolutionary Computation[C].Brisbane,Australia:IEEE,2012.1-8.
[4] Jin Y,Sendhoff B.A systems approach to evolutionary multiobjective structural optimization and beyond[J].IEEE Computational Intelligence Magazine,2009,4(3):62-76.
[5] Schneider G,Schr dl W,Wallukat G,et al.Peptide design by artificial neural networks and computer-based evolutionary search[J].Proceedings of the National Academy of Sciences of the United States of America,1998,95(21):12179-12184.
[6] Venske S M,Gon ξ alves R A,Benelli E M,et al.ADEMO/D:an Adaptive Differential Evolution for Protein Structure Prediction Problem[J].Expert Systems with Applications,2016,56:209-226.
[7] Zhang Q,Liu W,Tsang E,et al.Expensive multiobjective optimization by MOEA/D with Gaussian process model[J].IEEE Transactions on Evolutionary Computation,2010,14(3):456-474.
[8] Liu B,Koziel S,Zhang Q.A multi-Fidelity surrogate-model-assisted evolutionary algorithm for computationally expensive optimization problems[J].Journal of Computational Science,2015,12:28-37.
[9] Liu B,Zhang Q,Gielen G G E.A Gaussian process surrogate model assisted evolutionary algorithm for medium scale expensive optimization problems[J].IEEE Transactions on Evolutionary Computation,2014,18(2):180-192.
[10] Datta R,Regis R G.A surrogate-assisted evolution strategy for constrained multi-objective optimization[J].Expert Systems with Applications,2016,57:270-284.
[11] Jin Y,Sendhoff B.Reduction of fitness evaluations using clustering techniques and neural network ensembles:US,US7363281[P].2008.
[12] Pavelski L M,Delgado M R,Almeida C P D,et al.Extreme learning surrogate models in multi-objective optimization based on decomposition[J].Neurocomputing,2016,180:55-67.
[13] Montemayor-García G,Toscano-Pulido G.A study of surrogate models for their use in multiobjective evolutionary algorithms[A].International Conference on Electrical Engineering Computing Science and Automatic Control[C].Jilin,China:IEEE,2011.1-6.
[14] Deb K,Kalyanmoy D.Multiobjective Optimization Using Evolutionary Algorithms[M].Chichester:John Wiley & Sons,Inc,2001.
[15] Myers J L,Well A.Research Design and Statistical Analysis[M].Mahwah:Lawrence Erlbaum Associates Publishers,2013.
[16] Zitzler E,Deb K,Thiele L.Comparison of multiobjective evolutionary algorithms:Empirical results[J].Evolutionary computation,2000,8(2):173-195.
[17] Kursawe F.A Variant of Evolution Strategies for Vector Optimization[M].Heidelberg:Springer,1991.
[18] K Deb,L Thiele,M Laumanns,et al.Scalable multi-objective optimization test problems[A].2002 IEEE Congress on Evolutionary Computation[C].Honolulu,Hawaii:IEEE,2002.825-830.
[19] 郭观七,尹呈,曾文静,等.基于等价分量交叉相似性的 Pareto 支配性预测.自动化学报,2014,40(1):33-40. GUO Guan-qi,YIN Cheng,ZENG Wen-jing,et al.Prediction of Pareto dominance by cross similarity of equivalent components[J].Acta Automatica Sinica,2014,40(1):33-40.(in Chinese)
[20] Kleijnen J P C.Kriging meta modeling in simulation:A review[J].European Journal of Operational Research,2009,192(3):707-716.
[21] Rosales-Perez A,Coello C A C,Gonzalez J A,et al.A hybrid surrogate-based approach for evolutionary multi-objective optimization[A].2013 IEEE Congress on Evolutionary Computation[C].Cancun,Mexico:IEEE,2013.2548-2555.
[22] Deb K,Pratap A,Agarwal S,et al.A fast and elitist multiobjective genetic algorithm:NSGA-Ⅱ[J].IEEE Transactions on Evolutionary Computation,2002,6(2):182-197.
[23] Zhang Q,Li H.MOEA/D:a multi-objective evolutionary algorithm based on decomposition[J].IEEE Transactions on Evolutionary Computation,2007,11(6):712-731.
[24] Coello C A C,Pulido G T,Lechuga M S.Handling multiple objectives with particle swarm optimization[J].IEEE Transactions on Evolutionary Computation,2004,8(3):256-279.
[25] 谢承旺,邹秀芬,夏学文,等.一种多策略融合的多目标粒子群优化算法[J].电子学报,2015,43(8):1538-1544. XIE Cheng-wang,ZOU Xiu-fen,XIA Xue-wen,et al.A multi-objective particle swarm optimization algorithm integrating multiply strategies[J].Acta Electronica Sinica,2015,43(8):1538-1544.(in Chinese)
[26] 王亚辉,吴金妹,贾晨辉.基于动态种群多策略差分进化模型的多目标进化算法[J].电子学报,2016,44(6):1472-1480. WANG Ya-hui,WU Jin-mei,JIA Chen-hui.Multi-objective evolutionary algorithm based on dynamic population multi-strategy differential models[J].Acta Electronica Sinica,2016,44(6):1472-1480.(in Chinese)

基于相关分析的多目标优化Pareto优劣性预测

Prediction of Pareto Dominance Based on Correlation Analysis

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	钟沛龙, 黎明, 何超, 陈昊. 基于SOM聚类和自适应算子选择的高维多目标进化算法[J]. 电子学报, 2022, 50(8): 1959-1974.
[2]	魏振春, 傅宇, 马仲军, 吕增威, 石雷, 张本宏. 带时间窗的无线可充电传感器网络多目标路径规划算法[J]. 电子学报, 2022, 50(8): 1819-1829.
[3]	汪禹宏, 张屹. 基于适应度指导交配限制策略的重组算子与多目标优化研究[J]. 电子学报, 2022, 50(3): 710-717.
[4]	张屹, 陆逸舟, 王帅, 陆曈曈. 基于多源交配选择策略的重组算子与多目标优化研究[J]. 电子学报, 2021, 49(9): 1754-1760.
[5]	王蔚龙, 李勇军, 赵尚弘, 辛宁, 赵海燕, 张泰江. 基于两阶段帕累托优化的卫星下行链路功率分配方法[J]. 电子学报, 2021, 49(6): 1101-1107.
[6]	王亚良, 倪晨迪, 金寿松. 基于多邻居结构的自适应元胞差分算法[J]. 电子学报, 2021, 49(3): 578-585.
[7]	罗智勇, 朱梓豪, 谢志强, 孙广路. 面向云计算的花朵差分授粉工作流多目标优化算法研究[J]. 电子学报, 2021, 49(3): 470-476.
[8]	张磊, 刘庆, 杨尚尚, 杨海鹏, 程凡, 马海平. 基于双编码的重叠社团检测多目标优化方法[J]. 电子学报, 2021, 49(11): 2101-2107.
[9]	崔玉亚, 张德干, 张婷, 杨鹏, 朱浩丽. 一种面向移动边缘计算的多用户细粒度任务卸载调度方法[J]. 电子学报, 2021, 49(11): 2202-2207.
[10]	刘冰洁, 毕晓君. 一种基于角度信息的约束高维多目标进化算法[J]. 电子学报, 2021, 49(11): 2208-2216.
[11]	张红斌, 蒋子良, 熊其鹏, 武晋鹏, 邬任重, 袁天, 姬东鸿. 基于改进的有效区域基因选择与跨模态语义挖掘的图像属性标注[J]. 电子学报, 2020, 48(4): 790-799.
[12]	王海东, 刘云敬, 康琳, 武迎春. 射频能量捕获传感网移动能量源的布置策略研究[J]. 电子学报, 2020, 48(12): 2367-2375.
[13]	黎明, 黄珊, 陈昊, 李军华. 基于类圆映射的高维多目标可视化方法[J]. 电子学报, 2019, 47(6): 1185-1193.
[14]	谢承旺, 张飞龙, 陆建波, 肖驰, 龙广林. 一种多策略协同的多目标萤火虫算法[J]. 电子学报, 2019, 47(11): 2359-2367.
[15]	张磊, 毕晓君, 王艳娇. 基于重新匹配策略的ε约束多目标分解优化算法[J]. 电子学报, 2018, 46(5): 1032-1040.