The Research And Application Of Web Service Composition Based On Time-varying Petri Net

The technology of service composition can convert distributed services with single functions into a unified structure,which integrated complex applications.It's an important research branch in the field of service computing.Petri net contains the elements such as arcs,positions and transitions,which can describe the asynchronous,concurrent physical system properties.The theory of Petri Net can describe a business process and express the relationship between the function and operation,also detail the characterization of developing process of system status.The method of Petri Net modeling is based on graphical mathematics and has crucial research significance and broad applicate prospects.First,this paper aims at the timing quality parameters of service composing process.It's based on the random variables of QoS non-functional properties,and puts forward a kind of calculate method to comprehensive benefit values of QoS with the process of normalized and quantitative.With confirming the functional requirements of users,the calculate process includes the steps of statistic control structure and logical rules,dispose QoS data with classification processing.Then with the Petri net theory,this paper increases time varying functions to express the arcs in the service model of time varying Petri net,furthermore,it also describes the functions and logical structures.While services running in a specific period,the change process of system state is recorded by variable timing functions.Through the mapping rules,the model can construct the corresponding relations between functions and services performance,and abstract the combination into the collection of data and logic.In this paper,we also introduce the back path theory to describe the priority of scheduling mode,use the path testing to reduce the redundancy of services and improve the efficiency of service choice.According to the component resources in service community,we add a workflow correctness testing algorithm based on semantic,under the premise of users' requirements,we employ the artificial fish algorithm to obtain the optimal performance of composition,make sure to output the services with high efficiency and low cost.Finally,the paper designs a power plant information service system based on the scheduling rules.With the application of proposed model and calculation method,we obtain a series of test data from the actual electric power information service resources,and intuitively get the satisfaction and feedback results in interactive services users.Using the monitoring and control software to record service data,this system mainly implements the management service modules of code,drawing files and print services,which can optimize the actual data by combining the calculate method to service performance.Compared with classical service models,this paper analyses the indicators of reliability and availability of combination service,and conclude the proposed model with theoretical feasibility and practical effectiveness to describe time quality QoS and the comprehensive efficiency.