Multi-task Service Composition And Optimization Model Under Cloud Manufacturing Environment

Today,with the rapid development of information technology and the diversification of market demand,as a new business model,cloud manufacturing has always been a research hotspot in the manufacturing field.Based on modern information technologies such as cloud computing and the Internet of Things,cloud manufacturing can provide manufacturing resources or manufacturing capabilities in the form of services to users in a convenient ―pay-as-you-go‖ model.Users can select,configure and employ customized manufacturing services as needed.With the in-depth research in the field of cloud manufacturing and the gradual promotion of cloud manufacturing applications,more and more manufacturing resources are transformed into cloud manufacturing services for users to call.At the same time,the manufacturing tasks submitted by users are becoming more complex and need to be transformed into multiple smaller tasks.The entire manufacturing task is then completed in the form of a composition of cloud manufacturing services.Therefore,it is necessary to conduct in-depth research on multi-task service composition optimization under the cloud manufacturing environment.Cloud manufacturing task decomposition and cloud manufacturing resource function matching are the premise and foundation of multi-task cloud service composition optimization.However,the current research on cloud manufacturing mainly focuses on the cloud service composition optimization stage,but has less research on task decomposition and resource matching in the cloud manufacturing environment.In order to promote the application of cloud manufacturing,this thesis starts from the whole,and constructs a three-stage model including cloud manufacturing task decomposition stage,cloud manufacturing resource function matching stage and cloud manufacturing service composition optimization stage based on the analysis of cloud manufacturing tasks and cloud manufacturing resources,to deeply study the multi-task service composition optimization mechanism under the cloud manufacturing environment.The cloud manufacturing task decomposition stage decomposes the complex manufacturing tasks submitted by users into several subtasks of appropriate granularity.The cloud manufacturing resource function matching stage is responsible for matching each subtask to a manufacturing resource that can achieve its goal,forming a candidate cloud service set.The cloud manufacturing service composition optimization stage is to select a sutiable service from each candidate task cloud service set to form an optimal cloud manufacturing service composition based on the first two stages.The output of each stage is the input to the next stage.In the cloud manufacturing task decomposition stage,this thesis firstly divides the manufacturing tasks and decomposes the manufacturing process according to the characteristics of the manufactured products and manufacturing processes;Secondly,it is judged whether the granularity is appropriate by the degree of cohesion of the task;Finally,the decomposition of manufacturing tasks is completed according to the improved cloud manufacturing task decomposition rules,and the feasibility and effectiveness of the proposed improved task decomposition rules are illustrated by a specific example.In the cloud manufacturing resource function matching stage,this thesis firstly models and describes the ontology of cloud manufacturing tasks and cloud manufacturing resources;Secondly,it determines the functional matching elements of cloud manufacturing tasks and cloud manufacturing resources according to the concepts in the ontology model;Finally,an improved ontology semantic distance calculation method is proposed to determine the matching degree between ontology concepts,and the feasibility and effectiveness of the improved semantic distance calculation method are illustrated by a specific example.In the cloud manufacturing service composition optimization stage,this thesis firstly divides the general quality of service(Qo S)indexes of manufacturing cloud service into rigid indexes and flexible indexes,and the two categories of indexes are weighted by Entropy method;Secondly,the ELECTRE method is used to shortlist competitive manufacturing cloud service sets according to the rigid indexes;Thirdly,particle swarm optimization(PSO)algorithm is used to optimize manufacturing cloud service composition based on the flexible indexes,and the feasibility of the primary selection stage in the cloud manufacturing service composition optimization phase is illustrated by a specific example.At last,a specific case is used to explain the operation process of each stage and the relationship between stages,which verifies the feasibility and effectieness of the proposed model.