Research And Implementation On Compete Driven Allocation Method For Multi-Objective Optimization Of Cloud Manufactruing Resources

Cloud Manufacturing(CM)which is an advanced manufacturing mode combined with cloud computing and manufacturing technology attracted many scholars from all over the world once it was put forward.And it has a significant influence around the world.The resources allocation in CM is an essential part to match manufacturing capabilities with customer demands of cloud manufacturing platforms.And it plays a decisive role in improving the efficiency and stability of cloud manufacturing platform.Handling with a large number of tasks which have different objects and possibly competitive relationship in the cloud manufacturing platform,most of the current researched assume that manufacturing resources are unlimited.It is not exactly in line with the actual situation.In view of this,compete driven allocation method for cloud manufacturing resources was studied in this paper in order to meet the competitive demands of cloud manufacturing tasks for excellent resources and create an open,fair and efficient market environment for cloud manufacturing.Firstly,foundational theories and technologies of cloud manufacturing mode and resource allocation was discussed,the dimensions and structure of cloud manufacturing services was defined.After this,the problem of resources allocation in CM is described and defined.The overall implementation framework of online allocation for cloud manufacturing resources was proposed.Secondly,a framework for servitization of cloud manufacturing resources was proposed.Under the framework,the sensing and access methods for different types of resources are designed.Then the servitization of manufacturing resources are realized using JSON(Java Script Object Notation)and EAV(Entity-Attribute-Value model).The proposed solution of servitization improved the service level and controllability of physical manufacturing resources in cloud manufacturing system.In the same way,the publishing of tasks is also completed after f ormalization description of manufacturing tasks.On the basis of formalization of the manufacturing resources and tasks,the allocation of cloud manufacturing resources can be accomplished through three steps: coarse selection,matching and optimal decision.In consideration of the competitive relationship between tasks,the non-cooperative game theory was used to transform and solve the resources matching problem.Take time,cost and quality as three targets,multi objective equilibrium optimization model for revenue of multiple tasks was established.And the balance of efficiency and impartiality was achieved by finding Nash equilibrium point using fast non dominated genetic algorithm(NSGA-II).At last,the online allocation system for cloud manufacturing resources was designed and developed using tools of Java Web.The system holds the functions of publishing,managing and modeling for manufacturing resources and tasks.Which realized the online mathematical model establishment and solving of cloud manufacturing resources allocation,as well as the execution of resource allocation results performing by remote service invocation.A running example was demonstrated to verify the availability of the system.