| Cloud manufacturing is a new mode of networked and service-orientedmanufacturing inspired by the cloud computing. It aims to make manufacturingresource and manufacturing capability which widely spread in various enterprisesvirtualizes and encapsulate into manufacturing services. It is the research focuses ofadvanced manufacturing in recent years. Manufacturing cloud service composition isone of the key ways to reasonably configure manufacturing resources, finishmanufacturing task, and improve the manufacturing resource utilization rate.Manufacturing cloud service composition is divided into three stages, themanufacturing task decomposition, candidate manufacturing services discovery andoptimal-selection, manufacturing services composition and optimal-selection.Manufacturing service optimal-selection is the process to select the best manufacturingservices for manufacturing task. It is an important part of manufacturing cloud servicecomposition and its efficiency directly affects the whole service composition effect.Therefore, the research on manufacturing service optimal-selection has importantsignificance to implement manufacturing cloud service composition.At present, the research of manufacturing service optimal-selection mainlyconcentrated in two areas. Firstly, aiming to single resource service request for simpletasks, how to make the best choice from many candidates of manufacturing service.The existing research has not yet considered information and service quality dynamicchange and user feedback. Secondly, about multiple resources service requests forcomplex task, how to select one service from each sub-task’s candidate services andform a service composition in order to meet the task’s demand. The existing studiesonly consider a single level of service composition, in fact, complex tasks involvingmultiple levels of service composition. In view of the existing problems, around thecharacteristics of the simple tasks and complex task, this paper focus on optimal-selection method of manufacturing services. The chief content is as follow.Firstly, analysis the execution process of complex tasks, on the basis of theexisting research, summary and overview the overwhelming method of how todecompose the complex task into several simple tasks, how to carry out the process ofservice discovery and selecting best performance service for simple manufacturing task,and then, by service composition of the single function services, they accomplish complex tasks with cooperation. Point out the service optimal selection issue’srelationship in simple task and in complex task.Secondly, according to simple manufacturing task’s service optimal selectionproblem, divide service quality indicators into quantifiable indicators and those thatcannot be quantifiable, analyzed by using comprehensive weighted way andintuitionistic fuzzy set way respectively, put forward a new mechanism of valuation ofQoS involving users’ feedback.Finally, in view of the service optimal selection problems in complex task,proposed the hierarchical model of service composition, analysis the services qualitymodel, on the basis of the study above, use simulated annealing optimization algorithmto solve the problem, form a high quality of service method of manufacturing servicecomposition. Then, verify and analysis its performance in the Matlab simulationplatform, propose a feasible method for services optimal selection with complex task. |
PDF Full Text Request