A hybrid genetic algorithm approach for the "no-wait" flowshop scheduling problem

We describe a hybrid genetic algorithm (GA) approach to the no-wait flowshop scheduling problem using the minimize makespan as objective function. The GA approach described produces comparable or better results on benchmark flowshop problems than known heuristic methods. The representation used is a variant of one used on GA treatments of the travelling salesman problem (TSP). The genetic operators include edge recombination and three additional operators designed on analogy to known heuristic methods. Experiments show that these new operators are very effective in enhancing the performance of the traditional GA. The present approach also promises to effectively address the general job-shop problem.