Repair Methods for Box Constraints Revisited
Box constraints are possibly the simplest kind of constraints one could think of in real-valued optimization, because it is trivial to detect and repair any violation of them. But so far, the topic has only received marginal attention in the literature compared to the more general formulations, although it is a frequent use case. It is experimentally shown here that different repair methods can have a huge impact on the optimizer’s performance when using the covariance matrix self-adaptation evolution strategy (CMSA-ES). Also, two novel repair methods, specially designed for this algorithm, sometimes outperform the traditional ones.
Keywordsbox constraints repair method Baldwin Lamarck
Unable to display preview. Download preview PDF.
- 3.Suganthan, P.N., Hansen, N., Liang, J.J., Deb, K., Chen, Y.P., Auger, A., Tiwari, S.: Problem definitions and evaluation criteria for the CEC 2005 special session on real-parameter optimization. Technical report, Nanyang Technological University, Singapore (May 2005), http://www.ntu.edu.sg/home/EPNSugan
- 7.Purchla, M., Malanowski, M., Terlecki, P., Arabas, J.: Experimental comparison of repair methods for box constraints. In: Proceedings of the 7th National Conference on Evolutionary Computation and Global Optimization. Warsaw University of Technology Publishing House (2004)Google Scholar
- 8.Schwefel, H.P.: Evolution and Optimum Seeking. Wiley, New York (1995)Google Scholar
- 10.Eiben, A.E., Smith, J.E.: Introduction to Evolutionary Computing. Springer (2003)Google Scholar
- 11.Michalewicz, Z., Nazhiyath, G.: Genocop III: a co-evolutionary algorithm for numerical optimization problems with nonlinear constraints. In: IEEE International Conference on Evolutionary Computation, pp. 647–651 (1995)Google Scholar
- 13.Hansen, N., Auger, A., Finck, S., Ros, R.: Real-parameter black-box optimization benchmarking 2010: Experimental setup. Technical Report RR-7215, INRIA (March 2010)Google Scholar