| Zhengtong coal industry of Shandong energy Zikuang group is a first-class large-scale modern coal mine enterprise in China,During the construction of Gaojiapu mine,the enterprise has gradually established information systems related to production monitoring,such as industrial video system,personnel positioning system,geospatial management information system,etc.,but in the actual production process,the information and data of each link and each system are not shared and connected,resulting in the independent operation of each system,unable to interact with other systems,and not very good To support production decision-making.Based on the actual application requirements of Zhengtong coal industry,this paper designs and develops the Zhengtong industrial video big data management and analysis platform,which mainly completes the overall structure design,network architecture design,functional architecture design,workflow design,data warehouse architecture design,relational database design,class structure design,etc.;realizes the data resource management module,real-time video omni-directional display The functions of module,escape route management module,historical abnormal video management module,data modeling and analysis module,etc.enable the staff of Zhengtong coal industry to effectively manage industrial big data,and solve the problems of tedious and time-consuming traditional process of industrial big data management and analysis.The main work of this paper includes the following four aspects:First of all,according to the data source of Zhengtong coal enterprise,the platform is developed to further store the data through the data warehouse technology,collect the enterprise system log file,RDBMS data,HTTP extract data,local upload files and other data,so as to realize the real-time and effective management of data resources,as well as scientific and reasonable storage;Secondly,the governance method of data is proposed.Relying on Hadoop,RDBMS,disk array extraction and other technologies,the structured data and unstructured data in the production process of Zhengtong coal enterprise are managed efficiently,stably and rapidly,which greatly reduces the hardware cost,and solves the two problems of "can't save" and "can't find" industrial monitoring video in the Emergency Rescue Department of Zhengtong coal enterprise The core problem is to easily and quickly transfer out the monitoring video of the park to provide the first-hand video information for the mine safety detection.Thirdly,it analyzes the actual needs and business workflow of the employees in the data analysis department of Zhengtong company,and realizes the innovative call by simple drag and drop operation Spark,R algorithm package,such as Bayesian algorithm,decision tree algorithm,kmeans and other algorithms for data modeling analysis,high-performance server cluster calculation after the return of analysis results and PMML model,enabling Zhengtong coal enterprises to carry out in-depth mining and analysis of all kinds of data related to coal mine safety production,creating greater value.Finally,through the full application of the surveillance camera,the problem of data sharing between the surveillance system and other systems is solved.The real-time broadcast of industrial video and the analysis and collection of abnormal event video data are realized,including the association information with structured data such as personnel positioning system,geospatial management information system,emergency management system,etc.,and the generation of risk aversion on GIS map is developed At the same time,the escape route plays the core innovative applications such as real-time video along the way and historical abnormal video extraction,which further accelerates the rescue speed of the Emergency Rescue Department of Zhengtong coal industry enterprise,and improves the level of safety assurance during underground mining operation. |
PDF Full Text Request