Co-Evolutionary Algorithm For Fuzzy Flexible Job Shop Scheduling

Along with the deepening of the global economic integration and the accelerating change of user demands, the production methods of more variety and small batch have become more and more popular, which greatly increase the complexity of the production environment. How to reduce the unnecessary spending in the production process is very important for survival and development of corporation. Now corporations have the increasing urgent needs for more intelligent solutions to make production more economical. In the past 50 years, as a high economic benefits and np hard problem, production scheduling has been attracted broad attention by academics and industry.A large number of uncertainties and flexible factors make the manufacturing environment become complex. But in order to accurately describe the actual situation of the production process and improve the scheduling quality, we have to conduct a comprehensive and in-depth study. In this paper, the co-evolutionary algorithm is proposed for two kinds of scheduling problems in the fuzzy flexible job shop scheduling. The experimental results show that the proposed algorithm is great.The first problem is the fuzzy flexible job shop scheduling problem. First of all, we presented one kind of research methods called co-evolutionary methods after that the general research methods of flexible scheduling have been explore, the strengths and weaknesses of the consensus method and the decomposition method have been analysis. The co-evolutionary methods break down the problem into two populations and then use the fitness value make the two populations together, so that the two populations can affect the other one and co-evolution. In this problem, the first population is machine assignment; the second population is operation sequencing. An then, through a large number of simulation and comparative research, a co-evolutionary algorithm based on genetic algorithm and simulated annealing and kinds of optimization operators are designed. Finally, the experimental results show that the proposed algorithm has better performance than others by using three benchmark problems and comparing the solutions of the algorithm with different parameters.The second problem is the fuzzy flexible job shop scheduling problem with preventive maintenance. Considering the fact that preventive maintenance can reduce the probability of machine malfunction, we study the arrangements of maintenance period, put the handling of maintenance and decoding together, built an effective decoding process, proposed a two-population co-evolutionary algorithm. The experimental results of four examples show that the proposed algorithm is availability.