Tensor Theory-based Service Description Of Cloud Manufacturing Resources And Matching Modeling Research

Cloud manufacturing is a new modern intelligent manufacturing service model combining networked manufacturing and cloud computing technology.Through network communication and cloud computing service platform,the resources and capabilities of manufacturing resource providers are stored in cloud service providers through virtualization technology to provide efficient,intelligent and timely cloud manufacturing services on demand for requesters of manufacturing resources.Cloud manufacturing realizes centralized management of decentralized resources and decentralized services of centralized resources through utility computing,elastic computing,edge computing and other technologies.The technologies and theories related to materialization,virtualization,servitization,collaboration and intelligence are the main directions of research in the field of cloud manufacturing at present.In this paper,we mainly focus on the issues related to the key technology areas of cloud manufacturing resource servitization,with emphasis on the description,storage,and publication of manufacturing resources and capabilities as well as the search,matching,and combination of services involved in cloud manufacturing servitization.The main work of this paper:(1)Aiming at the problems of data heterogeneity and resource unification description in the process of decentralized service of centralized resources for intelligent manufacturing in cloud computing environment,a tensor theory-based encapsulation and publishing strategy for servitization of cloud resources is proposed by using the technical characteristics of tensor theory in describing heterogeneous data.Firstly,cloud manufacturing resources are divided into three categories: manufacturing resources,design resources and manufacturing capabilities;secondly,a tensor description encapsulation model for different types of cloud manufacturing resources is defined based on tensor theory,and low-order sub-tensors are established according to different resource service attributes,and then tensor extension operators are used to fuse the low-order sub-tensors,embedding various types of attribute information into the high-order tensor space to achieve a unified representation of cloud manufacturing resources;finally,a tensor description model for cloud manufacturing resources is given;Finally,the service-oriented deployment and publishing strategy of cloud manufacturing resources is given.(2)For the problem of matching manufacturing capability and service demand in cloud manufacturing environment,a multi-level cloud manufacturing capability demand intelligent search and matching method is given.First,the initial screening of the types of manufacturing capability services is carried out to filter some of the capability services at the release side,so as to obtain the service candidate sets that meet the types of production requirements,and then the basic attributes,functional attributes,state attributes,commercial attributes and quality attributes of the capability services are matched layer by layer to filter the set of capability services that meet the user’s task requirements for user selection;semantic matching,parameter matching and quality The matching problem of different types of data in requirement information is solved by using semantic matching,parameter matching and quality matching strategies.(3)Based on tensor theory,a prototype system of cloud manufacturing resource description and capability requirement matching is developed under the premise of unified encapsulation of service and requirement information.In the research of this paper,the problem of unified resource description with heterogeneous data in the process of servitization of cloud manufacturing resources is solved by introducing tensor theory,and the problem of matching demand search after service release is investigated,and finally the effectiveness of the proposed method is verified in the developed platform,which provides a new feasible idea for resource sharing in cloud manufacturing servitization.