Research And Application Of Service Composition And Integration For Typical Industrial Equipment Detection And Monitoring

"Research on cloud service technology of typical industrial equipment and product inspection and monitoring",as a key project in the national key research and development project,is a generic technology research based on national quality.It is imperative to build the demonstration application of the typical industrial equipment detection and monitoring platform and to promote it in the special inspection institutions in China.The detection and monitoring service for typical industrial equipment is developed based on the service-oriented architecture.With the rapid growth of the number of detection and monitoring services in the industry,the service quality level is not uniform and the solutions of service combination are numerous.On the other hand,with the continuous iteration of the typical industrial equipment detection and monitoring system,the invocation and interdependence between services are increasingly complex and the system functions are constantly expanding.At the same time,after system deployment clustering,the normal operation of the system puts forward higher requirements for load balancing and cluster fault tolerance.Therefore,how to select the appropriate detection and monitoring services among the many detection and monitoring services,and which architectural pattern should be adopted to integrate these services has become the focus of this paper.The specific research work is as follows:(1)This paper analysis and use QoS model based on mixed weight setting for the service selection process to sort and filter the services in the same abstract category and then select the local optimum.The QoS model based on the mixed weight setting not only considers the subjective preference of users but also takes the objective distribution characteristics of QoS index data as the objective factor affecting the weight when setting the weight of QoS indexes.(2)After building candidate service sets through service selection,this paper transforms the multi-objective optimization problem in service composition into a single-objective optimization problem by constructing its own multi-objective optimization model and objective function.(3)This paper proposes an improved cuckoo algorithm to solve the optimization problem of service composition.Because the initial population distribution of the original cuckoo algorithm is not uniform,which leads to its low population richness,this article uses the slitting random method to initialize the population to make the population distribution more uniform and richer.Secondly,the step size of levy flight in the original cuckoo algorithm is completely random,which makes the iterative process unstable.In this paper,the adaptive step length is proposed to replace the original step length,which makes the iterative process of cuckoo algorithm more stable when searching for the global optimal solution.(4)In terms of service integration,this paper selects Dubbo + Zookeeper architecture for service integration,and completes the construction of service integration framework and the integration of various services in the system based on the service integration requirements of typical industrial equipment detection and monitoring system.