Research On Resource Service Model And Resource Scheduling For Cyber-Physical System

Cyber-Physical System(CPS)is a system which combines calculation process with control process and communication process.CPS fuses information resources and physical resources to change the interaction between information world and physical world.Compared with WEB services,CPS has a lot of repetitive physical entity,every one of the physical entity provides the corresponding function,locates at different physical environment and has different performance(QoS).Because of repeated physical entity,when the task invokes resources,there are many resource scheduling schemes existing.In this thesis,the main idea is using service-oriented architecture to study how to build a service-oriented CPS architecture.Firstly,this thesis discusses the service-or:iented CPS layer architecture.On this basis,this thesis studies the differences and similarities between the CPS and WEB services and determines the feasibility of studying service-oriented CPS resource service model.Secondly,this thesis puts forward the concept of virtual resources and virtual physical entity to take place of the real resources and physical entity which have the same function,the same physical environmental information,different QoS of physical entities.By reference to the WEB service model,this thesis proposes the hybrid resource service model based on OWL and XML,this model uses OWL to describe the virtual physical entity,services provided by a physical entity and CPS task,uses XML to describe the QoS of a real physical entity.Then,on the basis of the CPS resource service model,this thesis proposes the task-virtual resource scheduling mechanism of CPS based on intelligent planning,this mechanism mainly determines to meet mission requirements of multiple candidate resource sets,this is the initial candidate sets of resources scheduling.Finally,according to the QoS of resources,this thesis proposes the mathematical model of resource selection for QoS based on multi-objection linear programming and uses the resource selection strategy based on multi-objection genetic algorithm to solve this model.The optimal solution meets the whole QoS optimal resource scheduling sequence in the resource candidate concentration.By construction the forest fire prevention system,this thesis has verified the effectiveness of the task-virtual resource scheduling mechanism of CPS based on intelligent planning;through the data instance,this thesis has built the mathematical model of CPS resource selection based on QoS,has implemented the resource selection strategy based on multi-objection genetic algorithm to solve this model,in order to verify the mathematical model is proposed in this thesis can be solution,the result is to meet the whole QoS optimal resource scheduling sequence;through the experiment has proved the proposed CPS resource service model and the resource scheduling mechanism has a guiding significance and reference value for the service-oriented CPS architecture.