DOI: 10.3724/SP.J.1219.2013.00677

Information and Control (信息与控制) 2013/42:6 PP.677-685

Comparison and Application of the Intelligent Recombination Operators in Multi-objective Evolutionary Algorithms

In order to improve the search capability of evolutionary operators in multi-objective evolutionary algorithms, the performances of the simulated binary crossover (SBX) operator, the self-adaptive SBX operator and the composite differential evolution operator in multi-objective evolutionary algorithms are compared in aspects of the rate and extent of approaching the optimal frontier through a set of experiments. A new composite evolutionary operator, named CSDO (Composite SA-SBX and Differential evolutionary Operator), is proposed, which integrates the modified directional SBX operator with the intensive hopping differential operator. The proposed operator is an efficient guarantee to the diversity of offspring and the speed of searching optimal individual. Meanwhile, a variable number of offspring archive technique based on the greedy strategy is proposed, which can improve the evolutionary rate. The technique enables more efficient use of the generated solution when considering the reserve value of optimum individual in whole population. The proposed CSDO operator and the modified archive method are both applied to test the benchmark problems, and simulation results indicate that the solution precision and the searching speed achieved using CSDO is superior to other intelligent recombination operators.

Key words:multi-objective optimization,multi-objective evolutionary algorithm,intelligent recombination operator,composite evolutionary operator,archive strategy

ReleaseDate:2015-04-15 18:52:37

[1] Coello C A C, Lamout G B. Evolutionary algorithms for solving multi-objective problems[M]. Berlin, Germany: Springer-Verlag, 2007.

[2] 徐波,彭志平. 一种基于云模型的多目标进化算法[J]. 信息与控制,2012,41(3):326-332. Xu B, Peng Z P. A multi-objective evolutionary algorithm based on cloud model[J]. Information and Control, 2012, 41(3): 326-332.

[3] Garza F M, Toscano P G, Coello C C. Alternative fitness assignment methods for many-objective optimization problems[J]. Artificial Evolution, 2010, 5975(10): 146-157.

[4] Batista L S, Campelo F, Guimaraes F G, et al. A comparison of dominance criteria in many-objective optimization problems[C]//IEEE Congress on Evolutionary Computation (CEC). Piscataway, NJ, USA: IEEE, 2011: 2359-2366.

[5] Deb K, Joshi A D. A computationally efficient evolutionary algorithm for real-parameter optimization[J]. Evolutionary Computation, 2002, 10(4): 371-395.

[6] Eshelman L J, Schaffer J D. Real-coded genetic algorithms and interval-schemeta[J]. Foundations of Genetic Algorithms, 1993, 2(3): 187-202.

[7] Ono I, Kobayashi S. A real-coded genetic algorithm for function optimization using unimodal normal distribution crossover[C]//Proceedings of the Seventh International Conference on Genetic Algorithms(ICGA-7). 1997: 246-253.

[8] Tsutsui S, Yamamura M, Higuchi T. Multi-parent recombination with simplex crossover in real coded genetic algorithms[C]//Genetic and Evolutionary Computation Conference (GECCO-99). 1999: 657-664.

[9] Voigt H M, Muhlenbein H, Cvetkovic D. Fuzzy recombination for the breeder genetic algorithm[C]//Sixth International Conference on Genetic Algorithms. San Francisco, California, USA: Morgan Kaufmann Publishers, 1995: 104-111.

[10] Deb K, Agrawal B R. Simulated binary crossover for continuous search space[J]. Complex Systems, 1994, 9(3): 115-148.

[11] Jain H, Deb K. Parent to mean-centric self-adaptation in SBX operator for real-parameter optimization[C]//The 2nd International Conference on Swarm, Evolutionary, and Memetic Computing, SEMCCO 2011. India: Andhra Pradesh, 2011: 299-306.

[12] Das S, Suganthan P N. Differential evolution: A survey of the state-of-the-art[J]. IEEE Transactions on Evolutionary Computation, 2011, 15(1): 4-31.

[13] Price K V. Differential evolution[M]. Berlin, Germany: Springer-Verlag, 2013: 187-214.

[14] Li H, Zhang Q F. Multiobjective optimization problems with complicated Pareto sets, MOEA/D and NSGA-II[J]. IEEE Transactions on Evolutionary Computation, 2009, 13(2): 284-302.

[15] Zitzler E, Thiele L. Multiobjective optimization using evolutionary algorithms - A comparative Study[C]//Parallel problem solving from nature - PPSN V. Berlin, Germany: Springer-Verlag, 1998: 292-301.

[16] Van Veldhuizen D A, Lamont G B. Evolutionary computation and convergence to a pareto front[C]//Late Breaking Papers at the Genetic Programming 1998 Conference. San Francisco, California, USA: Morgan Kaufmann Publishers, 1998: 221-228.

[17] Zitzler E, Thiele L. Multiobjective evolutionary algorithms: A comparative case study and the strength pareto approach[J]. IEEE Transactions on Evolutionary Computation, 1999, 3(4): 257-271.

[18] Deb K, Thiele L, Laumanns M, et al. Evolutionary multiobjective optimization[M]. Berlin, Germany: Springer-Verlag, 2005: 105-145.

[19] Deb K, Pratap A, Agarwal S. A fast and elitist multiobjective genetic algorithm: NSGA-II[J]. IEEE Transactions on Evolutionary Computation, 2002, 6(2): 182-197.

[20] Wang Y, Cai Z, Zhang Q F. Differential evolution with composite trial vector generation strategies and control parameters[J]. IEEE Transactions on Evolutionary Computation, 2011, 15(1): 55-66.