Copyright © 2006 Elsevier B.V. All rights reserved.
Synthesis of the β-distribution as an aid to stochastic global optimization
Available online 2 January 2007.
References and further reading may be available for this article. To view references and further reading you must purchase this article.
Abstract
The β-distribution is used as a point generation scheme in global optimization. Two population set-based global optimization algorithms are considered. These are the differential evolution (DE) and the controlled random search (CRS) algorithms. The point generation schemes of DE and CRS are hybridized with the β random variate. The hybridization uses a probabilistic combination of the point generation by the β-distribution and the point generation by DE or CRS. Numerical experiments are carried out using both test and practical problems. Numerical results suggest that the resulting algorithms are superior to their respective original counterpart.
Keywords: Global optimization; Population set; β-Distribution; Continuous variable; Probabilistic adaptation; Hybrid
Article Outline
- 1. Introduction
- 2. Brief descriptions of CRS and DE
- 2.1. The CRS algorithm
- 2.2. The DE algorithm
- 3. The self-adjusting β-density
- 3.1. Use of the β-distribution in CRS2 and DE
- 3.2. Implementation of β-distribution
- 4. Probabilistic combination of point generation schemes
- 5. Practical problem
- 6. Numerical results
- 6.1. Parameter value selection
- 6.2. Comparison
- 6.3. Comparison using PL function
- 6.4. Probabilistic adaptation
- 7. Conclusion
- Acknowledgements
- Appendix A. Proof of Lemma 1
- Appendix B. Proof of Lemma 2
- References






E-mail Article
Add to my Quick Links

Cited By in Scopus (0)






