Big Data Analysis And Application Of Pumped Storage Operation Monitoring System

With the development of SG186 and SG-ERP in State Grid,the reform of electric power and the development of new technologies such as "Big Data,Cloud Computing,Material Internet and Mobile Internet" in recent years,the importance of strengthen the application of big data in electric power enterprises,promoting the data value,constructing the data management system and enhancing the monitoring and analysis capabilities is much more obvious.Xinyuan Holding Company as the world's largest FM peak professional operating companies,it also hopes to use the insight to find capacity,process optimization capabilities and intelligent decision-making capabilities of big data in the era of large data.And then achieve the whole life cycle management that data assets from the generation,acquisition,storage,processing to the value mining,so as to improve the level of enterprise management and management quality,enhance the strategic decision-making ability of enterprises.This thesis is based on the internship experience in Xinyuan Holding Company monitoring center.Combining to the research in Xinyuan Holding Company,put forward the whole life cycle management concept of pumped storage enterprise operational data and asset management,and complete the following work:(1)Study the development situation and the existing achievements of Xinyuan Holding Company's main business at present,complete the demand analysis work for monitoring data assets management,and then put forward the whole life cycle management concept that suitable for pumped storage enterprise operational data and asset management.(2)Collected and compiled the production and operation data accumulated by the operation monitoring center,completed the pretreatment and cleaning of the large data of the production equipment with hadoop platform,and merged the operation data from all business system.(3)Based on the mass production operation data,we start the research on the operation intensity and reliability of the pumping equipment,and the operating states feature and the health status of the turbine units is researched secondly.Then,based on the ARMA and LS-SVM,the health status trend forecast model and the pendulum performance degradation forecasting model of the equipment operation is established.Based on this,the construction of equipment health level evaluation and fault analysis prediction system is completed.(4)The business data analysis and the index system construction are carried out for the enterprise operational data assets,further define the representative analysis of the theme,then construct data analysis model for the main business which needs key analysis and build a data analysis model respectively,also the data warehouses,finally expounds the process of data mining and data visualization display content.(5)Combined with BI visualization tools Tableau,set up a platform of panoramic visual data assets,put the business data assets analysis summary of mining integration on the platform,and carry out corresponding function test and performance test,thus to provide professional support for the promotion of professional management.