Design And Implementation Of OPC UA Distribute Message Middleware For WSN

In recent years,with the increasing level of technology in China,the "smart manufacturing" industry is developing rapidly.As one of the power sources of this industry,the technology model of "Internet + Industry" has effectively driven the transformation of the manufacturing industry to intelligence and digitalization,which involves key technologies such as Wireless Sensor Networks(WSN),TSN(TimeSensitive Networking)and OPC UA(OPC Unified Architecture).How to achieve the unification of wireless sensor networks and industrial Io T at the semantic level has now become a hot research topic in the field of "smart manufacturing".Wireless sensors are widely used in industrial sites to help improve industrial capacity utilization and automation.However,wireless sensor nodes themselves have limitations such as high system energy consumption and low storage capacity,and data communication between devices mainly uses OPC UA technology,which is traditionally designed for wired networks and uses the connection-based client/server(C/S)mode.Its direct application to wireless sensor networks has high resource overhead and low real-time performance.Therefore,this paper carries out research work on the design and implementation of OPC UA distributed messaging middleware for wireless sensor networks in conjunction with a small pilot production line.The contributions and innovations of this paper are as follows.(1)A wireless sensor network information model based on OPC UA is developed to address the problem of traditional communication methods that make it difficult to efficiently access sensing data to the industrial Io T.By creating nodes and assigning references in XML files,the wireless sensor network topology is dynamically mapped into the OPC UA address space.The test results show that the model can unify the information collected by wireless sensor nodes from different manufacturers with their own information,and facilitate the provision of a unified data access interface to data application systems.(2)To address the problem that the traditional OPC UA C/S model leads to a tightly coupled communication system and cannot realize one-to-many communication,an AMQP-based OPC UA distributed communication method is proposed,and an OPC UA publisher/subscriber mechanism is established.The data publisher and subscriber are decoupled by AMQP message agent.In other words,the publisher pushes the message to the switch of the proxy,and then delivers it to the public message queue according to the routing key,and the subscriber listens to the specified queue to receive the message.This mechanism effectively reduces the interaction of messages between communication objects.(3)To address the problem that it is difficult to define the abnormal boundary of data during transmission,an abnormal data detection algorithm H-DBSCAN for an intelligent production line is proposed.the algorithm addresses the problem of the high sensitivity of input parameters of the traditional clustering algorithm,and proposes a clustering parameter seeking mechanism based on the K-nearest neighbor algorithm;to address the problem that traditional clustering algorithm is greatly affected by the shape of data samples,and proposes a competitive idea-based the multi-cluster clustering merging mechanism is proposed.Tests on public data sets verify that the algorithm improves the accuracy of clustering and reduces the sensitivity of input parameters,thus improving the accuracy and intelligence of determining the abnormal status of devices.This paper develops an OPC UA distributed messaging middleware system based on PyQt5 based on the above innovation points,which mainly includes the data acquisition module,address space module,distributed communication module,and data detection module.Finally,based on the small experimental production line platform in the laboratory,the wireless sensor network is built and the system is tested for functionality and performance.The test results show that the system can meet the expected requirements and solve the existing problems of the intelligent production line described above to a certain extent,which not only can realize the distributed distribution of sensor data,reduce the network power consumption and improve the utilization rate of messages;but also can accurately mine the abnormal data and provide a reliable basis for the determination of the health status of the equipment.