A Proposal to Hybridize Multi-Objective Evolutionary Algorithms with Non-gradient Mathematical Programming Techniques
The hybridization of multi-objective evolutionary algorithms (MOEAs) with mathematical programming techniques has gained increasing popularity in the specialized literature in the last few years. However, such hybrids normally rely on the use of gradients and, therefore, normally consume a high number of extra objective function evaluations in order to estimate the gradient information required. The use of direct (nonlinear) optimization techniques has been, however, less common in the specialized literature, although several hybrids of this sort have been proposed for single-objective evolutionary algorithms. This paper proposes a hybridization between a well-known MOEA (the NSGA-II) and two direct search methods (Nelder and Mead’s method and the golden section algorithm). The aim of the proposed approach is to combine the global search mechanisms of the evolutionary algorithm with the local search mechanisms provided by the aforementioned mathematical programming techniques, such that a more efficient (i.e., with a lower number of objective function evaluations) approach can be produced.
KeywordsPareto Front Multiobjective Optimization Golden Section Nondominated Solution Multiobjective Evolutionary Algorithm
Unable to display preview. Download preview PDF.
- 9.van der Corput, J.G.: Verteilungsfunktionen. Akademie van Wetenschappen 38, 813–821 (1935)Google Scholar
- 17.Deb, K., Thiele, L., Laumanns, M., Zitzler, E.: Scalable Test Problems for Evolutionary Multiobjective Optimization. In: Abraham, A., Jain, L., Goldberg, R. (eds.) Evolutionary Multiobjective Optimization. Theoretical Advances and Applications, pp. 105–145. Springer, Heidelberg (2005)CrossRefGoogle Scholar
- 18.Veldhuizen, D.A.V.: Multiobjective Evolutionary Algorithms: Classifications, Analyses, and New Innovations. PhD thesis, Department of Electrical and Computer Engineering. Graduate School of Engineering. Air Force Institute of Technology, Wright-Patterson AFB, Ohio (1999)Google Scholar
- 19.Schott, J.R.: Fault Tolerant Design Using Single and Multicriteria Genetic Algorithm Optimization. Master’s thesis, Department of Aeronautics and Astronautics. Massachusetts Institute of Technology, Cambridge, Massachusetts (1995)Google Scholar