Font Size: a A A

Research On Fault Diagnosis Method In Transformer Remote Monitoring System

Posted on:2017-07-08Degree:MasterType:Thesis
Country:ChinaCandidate:J DuFull Text:PDF
GTID:2348330536476736Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Distribution transformer is the key equipment in the process of power transmission,and its running state has great influence on the whole distribution network.Finding the fault in transformer as soon as possible,and then giving suitable maintenance such as overhaul,debugging and so on,can effectively prolong the service life of equipment,increase utilization rate of equipment,and reduce loss of power distribution fault.The transformer remote monitoring system using dissolved gas analysis method can remotely monitor the status of transformer with no power off,along with continuous monitoring,anti electromagnetic interference,and many other features.It has an obvious advantage in monitoring and fault diagnosis,and the fault diagnosis method is the core part of the system.Firstly,after compared and analyzed the different multi pattern recognition methods of relevance vector machine,the multi-class relevance vector machine diagnosis model is constructed in this thesis.After training,this model can identify 7 kinds of fault states and also normal state of transformer,and obtains a good result in the case of limited sample data.Secondly,on the basis of the multi-class relevance vector machine diagnosis,for the different needs of fault diagnosis in monitoring system such as detailed fault types and visual display,this method is integrated with the improved three-ratio method and also the triangle graphic method,realizing different diagnosis methods.For the demands of both detailed fault types and visual display,these three methods is integrated together,adding an integrated diagnosis strategy,and improved visualization,which can not only improve the accuracy but also meet the different needs.After comparing the results of the integrated diagnosis method with each method separately,the result shows that this integrated method is superior to each single method.Lastly,with the locale Oil-Chromatographic equipment getting DGA data and transformer operating parameters,wireless remote device transmitting data,a transformer remotemonitoring system is designed and implemented,in which the integrated diagnosis method is used in fault diagnosis module.
Keywords/Search Tags:transformer remote monitoring, fault diagnosis, the multi-class relevance vector machine, integrated diagnosis
PDF Full Text Request
Related items