The effect of various operators on the genetic search for large scheduling problems

Genetic algorithms (GAs) have been applied on a variety of complex combinatorial optimization problems with high success. However, in relation to other classes of combinatorial problems, there is little reported experimental work for the application of GAs on large scheduling problems. The performance of a GA depends very much on the selection of the proper genetic operators. Crossover and mutation are the two major variation operators in any GA. This paper investigates the impact of various genetic operators on the genetic search through computational experiments carried out on the flow-shop scheduling problem (FSSP). A set of five crossover and six mutation operators are included in the experiments and their effectiveness on the overall performance of the GA process is measured, compared, and discussed. Furthermore, the case of crossover combination is examined under the FSSP framework investigating whether or not the various combinations outperform the sole usage of the best type of crossover operator.