Research On System Modeling And Scheduling Of Manufacturing Execution System Of Discrete Manufacuturing Industry

In flexible manufacturing environment, the production mode of discrete manufacturing industry makes the information during the production operation process more complex and harder to be controlled. MES (Manufacturing Execution System) is the effective means to communicate the information island between upper plan management and lower control, and how to manage production and scheduling effectively becomes critical. First, MES integration framework of discrete manufacturing industy is researched, which represents distributed, reconfigurable, integrable and scalable character, and the core scheduling problems is studied too. What’s more, static scheduling and dynamic scheduling are respectively studied and discussed in this dissertation. The main content of this paper includes the following aspects:Fistly, in view of difficult scheduling which is caused by complex production structure and process planning, the hybrid-hierarchical-control MES frame structure of discrete manufacturing industry based on Multi-Agent is put forward, information characters of real-time dynamic scheduling is taken into consideration too. The funcition, behaviors and interaction mechanism of each Agents, communication mechanism, methods and contents among Agents are studied. Furthermore, Multi-Agent cooperation method, interactive behaviors and task allocation are discussed based on contract net. So engineering needs and process of information transfer of discrete manufacturing industry is placed at flexible and effective interactive framwork.Secondly, according to the GA problems in solving process of the Flexible Job-shop Scheduling Problem (FJSP) now, an Adaptive Genetic Algorithm (AGA) is presented in this dissertation, and Job-shop scheduling problem (JSP) with partial flexibility request is transformed into a general FJSP. Moreover, a unified mathematical model is given. This model avoids many problems such as tedious encoding and decoding and no convergence caused by disunity of expression of description information and engineering models. Adaptive Genetic Algorithm (AGA) on the basis of iterative generation and AGA which adjusts corresponding parameters based on fitness values and distribution is respectively presented in this dissertation. Furthermore, a Total Adaptive Genetic Algorithm (TAGA) with both advantages of above two AGA is studied. Instance simulation verifies that the FJSP’s own characteristics are utilized in its solution by using AGA, which overcomes traditional GA’s limitation that initial stage of evolution is early and random search of medium-late stage is slow. Such methods are verified to accelerate convergence process, enhance searching efficiency and solving precision as well as avoid low efficiency and local optimum.Then, aiming at uncertain information and dynamic characteristic during job-shop scheduling process, some kind of dynamic scheduling method for FJSP is put forward based on real-time adjustment of controllable margin. A dynamic simuliaton solution mode framework is presented for FJSP. This framework is inspired by adaptive control, combined with the robust scheduling and foreseeing scheduling. It has both advantages of such two schedulings, and it’s stable and highly efficient. Preliminary scheme generation method based on foreseeing dynamics scheduling is introduced then. Foreseeing function is realized by fault-handling algorithm and dynamic simulation solver on the basis of AGA. Such project has consideration that random events may occur from the staticstic view. Thus, real optimistic results are obtained, and both the system adjustment and fluctuation of optimazation index are reduced when random events occuring. Strategy and methods of dynamic scheduling adjustment are brought forward, which respond to real-time event. In view of practical operation condition, two methods are proposed. One is rescheduling that condenses domain, the other is scheduling updating based on adjustment.Lastly, system modelling and scheduling of MES of discrete manufacturing industry researched in this dissertation are applied systematically. Directing at manufacturing management status and existing questions of reducer manufacturing factory, corresponding MES architecture is presented, prototype system is developed.