Bell-Curve Genetic Algorithm for Mixed Continuous and Discrete Optimization Problems
Rex K. Kincaid, Ruth Sykes, Michelle Griffith, Jaroslaw Sobieszczanski‐Sobieski · 43rd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference · 2002
This paper is the next installment in a series (Sobieszczanski-Sobieski et al. 1998, Kincaid et al. 2000, 2001, 2002 and Plassman and Sobieszczanski-Sobieski 2000) that has introduced a variant of the Genetic Algorithm in which the reproduction mechanism was modified to base it on the Gaussian probability distribution, the bell curve. The bell-curve based (BCB) heuristic procedure, first presented in Sobieszczanski-Sobieski, Laba, and Kincaid (1998), is similar in spirit to Evolutionary Search strategies (ESs) and Evolutionary Programming methods (EPs) but has fewer parameters to adjust. In Sobieszczanski-Sobieski et al. (1998) BCB was tested on a structural design optimization problem. The quality of solutions generated were verified by comparing BCB solutions to ones generated by a standard nonlinear programming technique. No attempt was made to analyze the sensitiv- The first author is the corresponding author. The first, second and third authors gratefully acknowledge the support of NASA-Langley Research Center--NAG-1-2077 ity of the BCB parameters. Kincaid, Weber and Sobieszczanski-Sobieski (2000) provide a preliminary investigation into BCB parameter selection as well as document improvements in the performance of BCB. Computational results for continuous, discrete and mixed continuous and discrete design optimization problems with constraints is reported. Further experiments with BCB for purely discrete optimization problems is provided in Kincaid, Weber and Sobieszczanski-Sobieski (2001). Plassman and Sobieszczanski-Sobieski (2000) implement and test a parallel version of BCB for continuous optimization problems