Robust Optimization Over Time: Problem Difficulties and Benchmark Problems

被引:36
作者
Fu, Haobo [1 ]
Sendhoff, Bernhard [2 ]
Tang, Ke [3 ]
Yao, Xin [1 ,3 ]
机构
[1] Univ Birmingham, Sch Comp Sci, Ctr Excellence Res Computat Intelligence & Applic, Birmingham B15 2TT, W Midlands, England
[2] Honda Res Inst Europe, D-63073 Offenbach, Germany
[3] Univ Sci & Technol China, Birmingham Joint Res Inst Intelligent Computat &, Sch Comp Sci & Technol, Hefei 230027, Peoples R China
基金
中国国家自然科学基金; 英国工程与自然科学研究理事会;
关键词
Benchmarking; dynamic optimization problems (DOPs); evolutionary algorithms (EAs); robust optimization over time (ROOT); DYNAMIC OPTIMIZATION; PARTICLE SWARM; ENVIRONMENTS; OPTIMA;
D O I
10.1109/TEVC.2014.2377125
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The focus of most research in evolutionary dynamic optimization has been tracking moving optimum (TMO). Yet, TMO does not capture all the characteristics of real-world dynamic optimization problems (DOPs), especially in situations where a solution's future fitness has to be considered. To account for a solution's future fitness explicitly, we propose to find robust solutions to DOPs, which are formulated as the robust optimization over time (ROOT) problem. In this paper we analyze two robustness definitions in ROOT and then develop two types of benchmark problems for the two robustness definitions in ROOT, respectively. The two types of benchmark problems are motivated by the inappropriateness of existing DOP benchmarks for the study of ROOT. Additionally, we evaluate four representative methods from the literature on our proposed ROOT benchmarks, in order to gain a better understanding of ROOT problems and their relationship to more popular TMO problems. The experimental results are analyzed, which show the strengths and weaknesses of different methods in solving ROOT problems with different dynamics. In particular, the real challenges of ROOT problems have been revealed for the first time by the experimental results on our proposed ROOT benchmarks.
引用
收藏
页码:731 / 745
页数:15
相关论文
共 37 条
[21]  
Li C., 2008, Benchmark generator for CEC'2009 competition on dynamic optimization
[22]  
Morrison R. W., 1859, P 1999 IEEE C EV COM, V3
[23]  
Nguyen T.T., 2011, Ph. D. Dissertation
[24]   Benchmarking and Solving Dynamic Constrained Problems [J].
Nguyen, Trung Thanh ;
Yao, Xin .
2009 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-5, 2009, :690-697
[25]  
Nuzzo P., P 2010 IEEE FORM MET, P71
[26]  
Ratschan S, 2006, ACM T COMPUT LOG, V7, P723, DOI 10.1145/1183278.1183282
[27]   Dynamic combinatorial optimisation problems: an analysis of the subset sum problem [J].
Rohlfshagen, Philipp ;
Yao, Xin .
SOFT COMPUTING, 2011, 15 (09) :1723-1734
[28]   Tracking moving optima using Kalman-based predictions [J].
Rossi, Claudio ;
Abderrahim, Mohamed ;
Diaz, Julio Cesar .
EVOLUTIONARY COMPUTATION, 2008, 16 (01) :1-30
[29]   Kalman-extended genetic algorithm for search in nonstationary environments with noisy fitness evaluations [J].
Stroud, PD .
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION, 2001, 5 (01) :66-77
[30]   Continuous dynamic problem generators for evolutionary algorithms [J].
Tinos, Renato ;
Yang, Shengxiang .
2007 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-10, PROCEEDINGS, 2007, :236-+