Improving Performance of GAs by Use of Selective Breading Evolutionary Process

In this paper, the use of selective breading evolutionary process for improving the performance of GAs is evaluated. To accomplish this evaluation, the generalized tardiness flow shop scheduling (GTFS) problem is designated. A natural evolutionary GA and two selective breeding Gas are developed for evaluating their performances in solving the proposed problem. An extensive numerical experiment on total of 2250 randomly generated scenarios is conducted to compare the effects of selective breeding mechanism. The effects of the varieties factors on the solution of the algorithms are analyzed by the factorial ANOVA. The computational results reveal that a significant improvement can be obtained if one employs an initial population with better genes.