Research On Cloud Manufacturing Service Composition And Optimization Method For Machining

With the new development trend of the global manufacturing under the influence of economic globalization,cloud manufacturing(CMfg)as a new manufacturing mode has become one of the important steps in the innovation of Chinese manufacturing industry towards information,service and network manufacturing industry.CMfg takes into account the idea of the combination of “centralized sharing of decentralized resources” and “decentralized service of centralized resources”.Machining process as on of the core links in manufacturing industry,has the characteristics of various manufacturing tasks,massive and heterogeneous manufacturing resources,complex specialized arrangement and collaboration mechanism and unbalanced manufacturing capacity distribution,leading to the coexist of bottleneck and idle of manufacturing capacity.Machining process has become one of the important research field in the field of CMfg application and development.How to allocate and composite manufacturing resources optimally is one of the important problems in the application and development of CMfg service model.Related research is hard to solve composition and optimization(CSCO)problems with acceptable space-time precision and efficiency.Existing solving method still have limitation in the accuracy and stability in solving CSCO problems.Therefore,this paper focused on the research of machining-oriented cloud manufacturing service composition and optimization and carried out corresponding research about above field in CMfg.Firstly,the machining manufacturing tasks and resources are analyzed.Based on the machining feature modeling approach and description CMfg model,a CMfg task requirement model and a manufacturing resource model are proposed.Additionally,the matching process of CMfg tasks and services are analyzed and service composition process is outlined,which has laid a solid foundation for next research steps.Secondly,CMfg single task CSCO problem is studied and a quality of service(QoS)evaluation method based on completion time,product quality and total cost is proposed,and the mathematical model of single CSCO problem is established in CMfg system.A hybrid chaotic artificial bee colony algorithm(ABCSA)with better solving ability is proposed by employing both simplex method and chaotic global best guided strategy.Simulation and analysis of experiments are carried out and the results clearly prove feasible and effective of ABCSA in solving single CSCO problems.Thirdly,as for multi-task CSCO with competitive resources in CMfg,on the basis of analyzing the difficulties and shortcomings of CSCO problems and scheduling problems,a multi-task service composition and optimization model considering resource competition constraints(MTSCOM-RCC)in CMfg environment is proposed.The MTSCOM-RCC model considers the interest and time conflicts between parallel multi-tasks’ subtasks and propose a new evaluation quality of service(Qo S).On the basis of the study of the MTSCOM-RCC model characteristics and mechanism,a two-layer coding scheme including task layer and service layer is proposed.An improved hybrid genetic artificial bee colony algorithm model(HABC)is proposed to solve the multi-task model.Experimental results prove that the two-layer task optimization method model is more feasible and effective in solving MTSCOM-RCC than the other two solution models.Finally,the prototype of CMfg platform is introduced.After analyzing the business process and the core module features,based on C# language and MYSQL database system,the core function are implemented and showed in pictures,including the tasks and service release,matching process,composition and optimization modules.