Modeling And Simulation Of Pull Job Scheduling System Based On Queuing Network

Job scheduling and optimization of manufacturing workshop is an indispensable part in production system optimization and control links, related to the continuity of production timely, directly affect the efficiency and effectiveness of production system. With the diversified and personalized of customer demands, multi-variety and small-batch production has become a trend, the job shop scheduling problem becomes more and more complicated.Pull-type control mode combine with the demands of customers and workshop site to complete the job scheduling, can compensate for the lack of push-type production mode effectively, so it easy to achieve rapid response to the market, reduce inventory and production cost. However, the pull-type control mode in the simple production line is running well, but in manufacturing various products, the application of technology is complicated, there are many problems to be solved. Therefore, the queuing simulation method and system based on network technology and through the push control, Kanban control, CONWIP control and Basestock Kanban CONWIP control under different modes of logistics operation optimization system modeling and simulation, all kinds of control modes were compared to explore the possibility and effectiveness of pull scheduling strategy mechanism in manufacturing practice application.In this thesis, firstly research multi-product and multi-stage manufacturing environment operation characteristics and the requirements of the scheduling system, on the existing push and pull production control strategy(PCS) has carried on the analysis and the summary, then put forward a Basestock Kanban CONWIP control strategy(BK-CONWIP), in order to make the system able to reduce WIP inventory and shorten the average production cycle, while maintaining capacity of flexibility, higher productivity and equipment utilization.For the further completion of the simulation system, this thesis uses queuing network theory established push control, Kanban control, CONWIP control and the proposed Basestock Kanban CONWIP control four kinds of modes of queuing network models respectively.On this basis, the thesis based on the scheduling problem of single-product and multi-product, simulation research based on ExtendSim platform, selects the manufacturingprocess permutation flow shop and the circuit board as the object, four modes of push control,Kanban control, CONWIP control and proposed the Basestock Kanban CONWIP control, and comparative analysis of the different control effects. For permutation flow shop scheduling problem of single product, further discusses the order fluctuations, rework fluctuations,fluctuations in the position of rework impact scheduling results; In view of multiple products circuit board manufacturing workshop, also research the deterministic demand and stochastic demand, the different control strategies are studied and compared.The simulation results show that the effectiveness of the pull-type control is higher than the push-type control; the proposed Basestock Kanban CONWIP control strategy in reducing WIP inventory, shorting the average production cycle time and improving the machine utilization rate is higher than the other three kinds of control strategies. In addition, it can reduce the impact of fluctuations in the production process, so as to ensure the system running smoothly.