The Research Of Multi-source Information Management System Of EMS Based On Data Stream Techniques

Due to the appearance of the phasor measurement unit(PMU) based on Global Position System(GPS),the synchronous phasor measurement will be applied widely in the power system.On account of technique and price,PMU can but be configured on the key nodes and cooperate with Supervisory Control and Data Acq uisition(SCADA) to form the mixed measurement system.The renovating frequency of PMU data in milliseconds will make the database in the energy management system(EMS) in existence can't deal with the large numbers of PMU data in time.Data stream technique as an emerging novel data management technology aims at the infinitude,speedy changing and persistent increasing of the data stream.As the PMU data and SCADA data are instantaneous and infinitude,the data stream technique adapt to manage them in the extreme.The paper researches the principle,method and key technology of managing the multi-source data of EMS by using the data stream techniques.Firstly,a design conceive of the multi-source information management system of EMS is proposed through analyzing the existing problem of the database in EMS in existence,and data stream and information stream are programmed.On the basis of the data stream model (DSM) matching all the data acquired in the EMS developed,the frame of the multi-source information management system of EMS based on the data stream techniques is proposed.Then,the main function module and the key technology in detail.Chapter 3,chapter 4 and chapter 5 describe the managing and storing process of the multi-source data stream and information stream of EMS.First,this paper proposes the notion of the diffluence of multi-source data stream.And then the accordant junction of the multi-source information stream is analyzed in order to form the mixed measurement system to realize the new application on the dynamic monitoring and stabilization calculation.At last,the paper researches the real-time and history information database and proposes the strategy of storing by sampling according to multiple time granularities. At the end of this thesis,the main results of the research are summarized.The further work is pointed out as well.