Study On Multi-objective Dynamic Scheduling Of Flexible Job-shop Based On Hybrid-driven Mechanism For The Degree Of Rescheduling Requirements

With the increasing development of market globalization, manufacturing enterprises nowadays face a lot of problems differently from the past during the increasingly fierce competition market environment. For example, the customers’ requirements are increasing, the number and types of products are increasing variously and complicatedly, the life of products is reducing gradually, and the imbalance between the environment and cost which caused by production is growing, etc. How to reduce the production cost, shorten the production period, ensure the production quality for the high competitiveness of manufacturing enterprises with limited resource have already become an important topic.Scheduling plays an important part in most manufacturing and production systems. The efficiency of scheduling directly affects the efficiency of enterprise production and cost control. In modern manufacturing mode, the uncertainties (change of customer’s need, change of manufacturing resource state, etc.) will cause the real-time change of workshop conditions. Especially in the Job Shop, a customer’s need-oriented manufacturing workshop, the uncertainties are much more, and the frequency and degree of change in working conditions are also much more than that of others. Therefore, the research of scheduling should close to the actual production demand and integrate as well as create more research results, in order to adapt to the precipitate changes during production and satisfy the real-time condition.This dissertation firstly introduces the development trend of modern manufacturing, the influence on production management, the basic concept and development trend of shop scheduling and the types, characteristics and research methods of scheduling. It summarizes the research status of MFJSP (Multi-objective Flexible Job-Shop Scheduling Problem) all over the world and points out the deficiency of existing research. Based on the analysis about the content and function of the job-shop scheduling problem, this dissertation provides the theoretical framework of MFJSP, and also summarizes and analyzes the research contents, crucial technology, related theory and technology about MFJSP.Then the dissertation raises the rescheduling driven mechanism under the in-depth analysis of disturbance classifying about flexible job-shop based on the degree of reschedule demand. The mechanism, which combines the ROC (rate of change) productions effected by the uncertainty during the production and production efficiency rate effected by the feasibility of rescheduling under real-time production condition forms the demand degree of production rescheduling. When the disturbance occurs, the first step is to calculate the ROC of working schedule and efficiency rate of working equipment. Then we should calculate degree of reschedule demand in order to decide whether to trigger the reschedule calculation.The next step is to raise the scheduling goal functions about how to minimize the makespan, minimize the production cost and minimize the maximum processing fraction defective based on the description of flexible job-shop scheduling problem. It also designs a multi-goals, optimized strategy based on multi-objective weighted optimization and Pareto non-dominated solution sets. The dissertation, which aims at the multi-goals optimized problem, raises a MFJSP solution algorithm based on blood variation genetic algorithm, and finally verifies the algorithm successfully using the comparison tests which based on real examples.Besides, the dissertation proposes the entirety concept model and the basic thinking of scheduling about real-time statue of flexible job-shop scheduling system based on the analysis and research about the theory and method of flexible job-shop scheduling. It also introduces the method about case-model and database model by using UML model and some interfaces of system prototype are given. It shows the visual, real-time, network and integration features to verify the validity of the proposed theory.Finally, the dissertation summarizes the whole research works appeared in it and looks forward the further research works about MFJSP.