| Fused magnesia is a kind of high temperature refractory material widely used in aerospace,chemical metallurgy,building materials and other fields,which is important strategic resource.The smelting process of fused magnesia is very complex,and the smelting process corresponds to a large number of production process data.With the acquisition technology development,the business of data collection is improving in enterprise.In addition to traditional relational sensor data such as current and voltage in the manufacturing process,the enterprise also enhances the collection of streaming media data like video,audio and operational logs.Therefore,the current production process data has the characteristics of complex structure and large value density.The rapid and effective analysis of a large amount of accumulated data and the excavation of the mechanism and process information which contained in the data have important guiding significance for enterprise to plan production reasonably,to realize intelligent production,save costs and improve efficiency.However,the traditional single data management system can only manage data of a single structure and cannot manage heterogeneous data of multiple structures and different types of data efficiently and uniformly.Moreover,different database systems in the data storage,data organization and other aspects bring huge time and hardware costs to resource management,relationship and behavior management.With the development of big data platform,the unstructured database is rapidly emerging.Therefore,this paper designs a distributed heterogeneous data management system based on distributed storage architecture.Based on realizing the traditional storage management function of structured data,heterogeneous data management system increases the storage,query and management of unstructured data.Starting from the actual needs of the factory and relying on process-oriented industrial big data management system research and development plans in the State Key Laboratory of Synthetical Automation for Process Industries at the Northeastern University,this paper carries out the related research.Aiming at the limitations of the traditional single data management system,a heterogeneous data management system for fused magnesium furnace(FMF)is designed and developed,which is used to manage the heterogeneous data of FMF efficiently.Meanwhile,the heterogeneous data management system is applied to a fused magnesia plant in Liaoning Province.The application results verify the effectiveness and practicability of the system.The main contents of this paper are as follows.Firstly,fused magnesia enterprise needs a heterogeneous data management system.Aiming at the problem of single function and data isolation of the existing data management system,combined with the Internet of Things heterogeneous data management system and based on the latest data management architecture,a heterogeneous data management system for fused magnesia furnace is designed.Its functions include data acquisition,data transmission,data storage,data query and data display.Secondly,aiming at the problem of the poor connection between each heterogeneous data(structured and unstructured),this paper designs a unified storage query solution for heterogeneous data.According to the actual demand of batch data query,the related parameters of data storage component are analyzed and adjusted to optimize the data query performance based on the query response time.Thirdly,the communication bandwidth of the enterprise is took into account.Based on the size of data packet in the process of data transmission of fused magnesia and the system delay and throughput as the performance indexes,the solutions of multiple data transmission proposed in this paper are compared and analyzed.Under the premise of meeting the data throughput,the data transmission scheme with the smallest delay is selected.At the same time,with the actual background,the related computing tasks are optimized by parallel computing,which can improve the execution efficiency of the communication tasks.Finally,the system developed in this paper is applied to a fused magnesia plant in Dashiqiao City,Liaoning Province,and its functions are verified for industrial application.The system mainly includes functions such as data acquisition,data transmission,data storage,data query and data display.At the same time,the system performance verification.The results of application verification show that the heterogeneous data management system developed by this paper meets the demand of efficient data management in industrial process and improves the level of data management. |
PDF Full Text Request