Research On Fuzzy Job-shop Scheduling Problem Based-on GA

With the coming forth of the global economy integration and knowledge economy, the competition between enterprises will be more drastic. In order to increase their core capacity of competition, enterprises must improve their inner production management, especially the core of production management-production scheduling. It is the key problem to make the full use of resources, work out the rational production plan, ensure the delivery just in time, lower the cost of production and heighten the efficiency of utilizing the production equipments. In the domain of production scheduling, job-shop scheduling problem is the most prevalent, complicated and difficult problem. Furthermore, because it can be popularized and transplanted, it has been paid close attention to by both academia and industry. Nevertheless, scholars often focus on certain job-shop scheduling problem in the past, but because it is affected by many uncertain factors in reality, this dissertation researches about fuzzy job-shop scheduling problem in depth based-on existing theories.In this dissertation, the current changing trend of circumstance in manufacturing is analyzed firstly. Several advanced manufacturing ideas are introduced simply. The research course and methods are reviewed. Furthermore, the definition of job-shop scheduling problem is given. Its characteristics are analyzed. The complexity of computing it is discussed. The mathematics model about it is constructed. And several important neighbor-region searching algorithms which are the most effective intelligent methods until now are discussed. The advantages and disadvantages of different coding, genetic operations are compared and analyzed. Then the dynamic adaptive GA is designed towards solving job-shop scheduling problem. The strategy of "hold best result" and "repeated crossover and mutation" are united into GA and the possibility of crossover and mutation can be adjusted automatically according to the results of optimization. The processing time of operation and duedate are represented by fuzzy numbers and the satisfaction degree of customers is presented for showing the degree that the completion time is satisfying with customers. Then using the operations and evaluation criterions of fuzzy numbers, multi-objective fuzzy job-shop scheduling problem is defined and studied.The designed GA is used as a method for solving fuzzy job-shop problem after it is validated that it is a very applicable and effective method by benchmark problems. The results of simulation experiments indicate that this GA can gain the best optimization results quickly and avoid the problem of "immature convergence" existing in the traditional GA. The results towards multi-objective fuzzy job-shop scheduling problem are beneficial for the realistic production in workshop. In the end, the whole dissertation is summarized and furtherdevelopments on fuzzy job-shop scheduling problem are explored.