Optimizing the performance of genetic algorithms for finding the optimal value of a given function

Abstract

An approach for optimizing the performance of genetic algorithms (GAs) which is derived from the exhaustive examinations of some parameters of GAs is provided. The problems of finding the optimal values of some numerical functions are used as examples to illustrate the performance of GAs. GAs are shown to be effective for solving these problems. In addition, various parameters of the optimization algorithm are critically selected for efficiency. Experimental results suggest that while it is possible to optimize GA control parameters, excellent performances can be obtained with an appropriately selected range of GA control parameter settings, based mainly on the experience of the users.<>