Research On Dynamic Scheduling Of Multi-variety Small Batch Production Mode Based On Hybrid Set Planning

With development of market economy, the typical flow-shop production pattern of large quantities becomes more and more difficult to adapt to the complicated and changeable market requirement currently. And the multi-type and small-batch production pattern is becoming the main production pattern of discrete manufacturing industry, with its flexible characteristic highly conforming to the demand of rapid market requirement respondence. In this production pattern, the products have a wide range of items and complex processes. What's more, the timely respondence to dynamic disturbances of the production process is also demanded. For these reasons, the scheduling of multi-type and small-batch production pattern is too complex. And only the dynamic scheduling method can solve it.Mixed Set Programming (MSP) is a practical method to solve the industrial complex combinatorial optimization problems of large scale. It has been applied in the aviation and railway transportation field and has solved some production scheduling problems. In this graduation thesis, MSP will be used to solve the dynamic scheduling problem of multi-type and small-batch production pattern. It is a further exploration and development of the MSP theory in the fields of production scheduling research and application.The main research results are as follow:(1) Firstly, a static scheduling problem solving model of multi-type and small-batch production pattern is built. Not considering the dynamic disturbances temporatily, a systematic analysis and summary of the multi-type and small-batch production pattern scheduling problem is done, with a result of building the constraints and optimization objectives. And then, the Natural Constraint Language (NCL) which is a supporting languge system of MSP is used to carry on the mathematical logic modeling of the constraints and optimization objectives. After that, an efficient algorithm is designed according to the optimization objectives. In this model, minimizing the total delay of jobs is the main optimization objective, and minimizing the total idle time of resources is the secondary optimization objective. Priotity constraints, assembly order constraints, optimal utilization of bottleneck resources and proportionality of resource utilization are taken into consideration, leading to a more optimized model and more reasonable results.(2) Secondly, the dynamic scheduling method of multi-type and small-batch production pattern is designed. Taking the dynamic disturbances and weekly planning of manufacturing shops into consideraition, a dynamic scheduling method of weekly planning and events driving to rescheduling is designed based on the dynamic scheduling pattern of periodical planning and events driving to rescheduling. This dynamic scheduling method is based on the static scheduling problem solving model, with adding dynamic order sets and dynamic constraints, leading to the functions of weekly planning and rescheduling in the conditions of rush orders coming, equipment failure and delivery date changing. Furthermore, a rolling scheduling method is designed to solve the problem that the operation time is too long when there are too many orders in the weekly planning. This rolling scheduling method can shorten the operation time greatly in the same conditions.(3) Finally, the dynamic scheduling method is verified by a job-shop production instance. It turns out that this dynamic scheduling method meets with effective and feasible results both in the weekly planning and rescheduling of events. The scheduling results provide scientific and quantitative decision bases for the dispatchers.