Research On Resources Selecting And Production Scheduling In Networked Manufacturing Floor Shop

Production scheduling is viewed as a key element to implement modernization production management. In this thesis, manufacturing resources selecting and the job-shop scheduling in MES are investigated in the networked manufacturing floor shop. And Multi-agent technology is applied to the modeling of scheduling system. Intensive study is made on the building scheduling model based on MAS, and genetic algorithm GA is also adopted in the algorithm for job-shop scheduling. This method makes scheduling system adapt the complex manufacturing environment, and improves the practicability and effectiveness of MES. Resources selecting is an important problem in the networked manufacturing floor shop. Firstly, in the thesis, an approach of manufacturing resources selecting is discussed, and the specialty of the fixture resources in the networked manufacturing is analyzed. AHP is investigated and used to realize manufacturing resources selecting on the platform in the context of ASP. Next, a scheduling framework and module based on multi-agent technology is constructed under the application background of distributed-manufacturing enterprise, which support the argumentation to design and implement dynamic scheduling in MES. A generic structure and module of manufacturing agent is presented. Four agent models for job-shop scheduling are presented. Besides, the mechanisms of communication and collaboration among these agents are probed. Then, format and content of message are described as well. Then, a hybrid genetic algorithm is proposed for NT-MES scheduling on the base of the scheduling framework of multi-agent system. Dynamic scheduling is achieved by interaction among these agents. Furthermore, some key elements of GA such as code and decode schema, crossover and mutation operators are investigated. Local search is applied to attempt to improve the solution after mutation. In addition, scheme and solution of prototype system are given. Subsequently, MES scheduling prototype system based on theories and methods as mentioned above is developed by using oriented object and software engineering technology, including of the functions of plan task management, process information management, simulation of production processes and work cell information. In the end, research of this thesis is summarized and research trend on scheduling in MES is anticipated.