Optimal Scheduling Of Cyber-Physical Production Systems

With the rapid development of today's science and technology and the changes in market demand patterns,manufacturing industry is transforming from extensive and centralized control mode to enhanced precision control mode which is decentralized.To meet these changes,the fourth industrial revolution begins gradually which is represented by the Germany's "Industrial 4.0" and led by intelligent manufacturing.This revolution achieves decentralized and enhanced manufacturing model through combining CPS(Cyber-Physical Systems)with advanced manufacturing technologies deeply to form CPPS(Cyber-Physical Production Systems).CPPS are distributed and heterogeneous,which achieve the changes in control mode from centralized to decentralized,and establish a highly flexible personalized and digital production model.The methods of developing the system operation strategy and handling the relationship between different units to make the system get the data for the tasks and systems effectively and dispatch different resources under different goals determine the quality and efficiency of the manufacturing system.Therefore,CPPS for discrete manufacturing is set as the research object in this paper and the scheduling strategies are studied.Firstly,the form of process knowledge base,the way to define task and device,and the way to match task and device are studied to provide basic data suporting for system operation.Secondly,the goals and constraints of system are set,and a two-layer decision model is designed based on dynamic decision cycle and center using MAS(Multi-Agent System).The optimization indicators,negotiation mechanism and method of re-scheduling are dissertated in detail to guarantee the scheduling optimization results.Thirdly,to make systems identify and obtain data for tasks and devices effectively and run following the schedule logic,the model of system operation logical based on CPN(Colored Petri Net)is built,including processing route CPN,execution CPN and scheduling CPN.Finally,the applicability and superiority of the strategies are proved through experiments and case studies.