The Job Scheduling System Research And Implementation Of The Discrete Manufacturing Workshop

As the market competition with dynamics, globalization and user driven becomes fiercer,and the complexity and dynamics of discrete manufacturing system increases, it makes higherrequests for production management of workshop. Job scheduling is one of the core contentof discrete manufacturing production management system, which is very important foroptimal use of its resources, improving product efficiency and cost savings. However, in theactual production process, the scheduling system out of touch with the actual productionenvironment, which leads the scheduling system can’t revise dispatching commands in timeaccording to the actual production environment，resulting in error accumulation.To solve the current problems of scheduling system, the thesis’s works mainly includesthe following four aspects:(1) For the problem of inaccurate process data and poor real-time information feedback,using RFID and Zigbee technology to realize the track of workshop manufacturing process,equipments state and workers state. By collecting the dynamic data and the real-time feedback,the scheduling system becomes a closed loop. Due to the Zigbee characteristics in workshop,it proposed a Grid and virtual force directed GA algorithm based energy efficient routingstrategy, making better performance in energy saving and prolonging the network productiontime.(2) According to the dynamic feature of discrete workshop, a physical job schedulingmodel based on Multi-Agent is established combined with the independence and coordinationcharacteristics of Multi-Agent technology. The analyzing scheduling system obtained thetarget function based on nine kinds of agents such an management agent, plan agent,inventory agent, task agent, process agent, data collection agent, equipment agent, employeeagent, schedule agent. The function and structure model of agents are studied. Furthermore,multi agent cooperation mechanisms are discussed base on Contract Net Protocol and theabnormal collaboration process are studied.(3) To solve the multi-objective problem in job shop scheduling considering the resourceconstraints of machines and operators, Fuzzy Pareto Dominance-Combined Biogeography-Based Optimization Scheduling Algorithm(FPDCBBO) is encapsulated in the schedule Agentto solve this problem, the design of the algorithm and its application in job shop scheduling aredescribed in more detail. At last, the algorithm can complete the job scheduling and resourcesallocations on the steady state and disturbance state of equipment failure by simulationexperiment,.(4) Finally, according to the actual demand of the mould workshop and the feature of theMulti-Agent job scheduling system, and along with the discussion of the system architecture,the functions module, the selection of data acquisition terminal and the database design, theMulti-Agent job scheduling is developed and implemented.