A parallel genetic algorithm for the job shop scheduling problem

This paper presents a parallel genetic algorithm for the job shop scheduling problem (JSP). There are following innovations in this new algorithm: active schedules are created by the priority rules of Giffler and Thompson [1]; the mutation uses neighborhood searching techniques; the crossover uses GT algorithm and is performed on 3 parents. We illustrate this new method on the parameters of Muth and Thompson's benchmark problems. It can produce optimal solutions at a high percentage of accuracy. Our proposed method is preeminent in comparison with other methods on both the calculation time and the speed of finding optimal solutions.