Research on an orthogonal and model based multi-objective genetic algorithm

ABSTRACT

Against low efficiency of traditional multi-objective evolutionary algorithms and poor utilization of Pareto-optimal solutions distribution regularity etc, in this paper, a new approach OMEA is proposed. It uses that distribution regularity to obtain good solutions, we also apply the orthogonal design to initialize population. Compared with SPEA2, NSGA-II and PAES, Pareto solutions by OMEA are closer to Pareto-optimal Front. The result of experiments shows a group of Pareto solutions with better convergence and diversity can be achieved, which gives strong supports to actual applications.

REFERENCES

• 1: Yaochu Jin and Bernhard Sendhoff. Connectedness, regularity and the success of local search in evolutionary multi-objective optimization. In Congress on Evolutionary Computation, pages 1910--1917, Canberra, Australia, December 2003. IEEE.
• 2: Aimin Zhou, Qingfu Zhang, Yaochu Jin, Edward Tsang, and Tatsuya Okabe. A model-based evolutionary algorithm for bi-objective optimization. In Congress on Evolutionary Computation, Edinburgh, U.K, September 2005. IEEE.
• 3: Aimin Zhou, Yaochu Jin, Qingfu Zhang, Bernhard Sendhoff, Edward Tsang. Combining Model-based and Genetics-based Offspring Generation for Multi-objective Optimization Using a Convergence Criterion{C} .cec2006.
• 4: Y. W. Leung and Y. Wang: An Orthogonal Genetic Algorithm with Quantization for Global Numerical Optimization. IEEE Transactions on Evolutionary Computation, 5(1) (2001) 41 -- 53.
• 5: Zhihua Cai, Winyin Gong, and Yongqin Huang, A Novel Differential Evolution Algorithm based on µ-domination and Orthogonal Design Method for Multiobjective Optimization, proceedings of the 4th International Conference on Evolutionary Multi-Criterion Optimization (EMO-07), LNCS 4403, Springer-Verlag. 2007.3.
• 6: Lindsay I Smith. A tutorial on Principal Components Analysis {EB/OL}. http://www.cs.otago.ac. nz/cosc453/student&tutorials/principal_components.pdf.
• 7: Deb. K, L. Thiele, M. Laumanns and E. Zitzler: Scalable Test Problems for Evolutionary Multi-Objective Optimization. KanGAL Report Number 2001001 (2001).
• 8: Deb, K., Pratap, A., Agarwal, S., Meyarivan, T.: A fast and elitist multiobjective genetic algorithm: NSGACII. IEEE Transactions on Evolutionary Computation 6 (2002) 182 -- 197.
• 9: Zhihua Cai, Wenyin Gong, and Yongqin Huang. A Novel Differential Evolution Algorithm based on µ-domination and Orthogonal Design Method for Multiobjective Optimization, proceedings of the 4th International Conference on Evolutionary Multi-Criterion Optimization (EMO'07), LNCS 4403, 2007.3, 286--301.

INDEX TERMS

Primary Classification:
  G. Mathematics of Computing
  G.1 NUMERICAL ANALYSIS
      G.1.6 Optimization
          Subjects: Global optimization

Additional Classification:
  F. Theory of Computation
  F.2 ANALYSIS OF ALGORITHMS AND PROBLEM COMPLEXITY
      F.2.1 Numerical Algorithms and Problems
          Subjects: Computations on polynomials

General Terms:
Algorithms, Experimentation, Performance

Keywords:
OMEA, multi-objective optimizing, orthogonal design, probability-model