Research On Intelligent Equipment Management System And Big Data Application In Power Plant

The power generation equipment is charactered with numerous systems and complex carvings,and the process parameters and the equipment running states require efficient and accurate monitoring and analysis continuously,which may have a profound and lasting impaction on reliability and efficiency simultaneously.Under the requirements of high standards of industrialization and informatization,scientific equipment management and data mining play a positive role in improving the production quality of the whole power plant.This thesis makes the following research on the interconnection of equipment information management system and SPSS Modeler:The thesis starts with a brief introduction of the structure and the theory of the server,the client PC and the Android PC.Under the consideration that the coupling of the system and SPSS Modeler is mainly based on the database connection,the database table and code is presented principally.For the client side,this design uses MySQL database,and SQLite database for Android side which has large data flow but small storage to adapt to the PC.Secondly,aiming at the huge scale and complicated types of the data,the thesis proposes the simple association rule analysis based on Apriori alghrithm.The scale and value of data in power plants are positively correlated with the high-tech,intelligent degree of equipment and the refinement of equipment management.The same goes for DM results and the integrity and historicity.For the purpose of equipment management,energy saving and consumption reduction,unit optimization,fault diagnosis,state analysis.Data mining is conducted in a particular sequence of data preparation,data cleaning,data selection and contraction,modeling and analysis,optimization of configuration,and differences analysis which shapes feedback loops.Finally,using the vibration monitoring data in Baiyin power plant as the sample,hiring the two-step cluster,anomaly analysis,feature selection,principal component analysis,Apriori and other modeling nodes of SPSS to analyze,and as a credible consequence,parameters are continuously reduced,and results are displayed in the form of association rules.The results are also visualized by the "mesh graph" function node.