On job-shop scheduling by a hybrid genetic algorithm

The rapid re-engineering of enterprises requires the shortest manufacturing time. Job-shop scheduling is one abstract model of production scheduling. In recent years, scheduling algorithms based on local search technology have received more attention. Methods such as simulated annealing, Tabu search, and genetic algorithms have been applied. In this paper, a production scheduling problem is discussed and an effective dynamic hybrid genetic algorithm is proposed to solve the complex F/T/1/1 problem successfully. In the research, an adaptive genetic algorithm is introduced and a new set of encoding rules and a fitness function are used. The new algorithm is analyzed in detail and the rules of repairing the bad chromosome are given in order to avoid the search failing because of such a bad chromosome. The simulation results show that the new algorithms and rules are suitable for production scheduling and the dynamic hybrid genetic algorithm is more effective than others.