Font Size: a A A

Study On Online Monitoring System Of Dissolved Gas In Transformer Oil

Posted on:2019-05-19Degree:MasterType:Thesis
Country:ChinaCandidate:L D MaFull Text:PDF
GTID:2382330545460140Subject:Engineering
Abstract/Summary:PDF Full Text Request
The transformer is the key equipment of the power system,its safe and stable operation directly relates to the safety of the power system.At present,transformers use mineral oil(transformer oil)as the insulating and cooling medium.During the operation,the insulation of the transformer is gradually deteriorated and the insulation capacity is reduced,the gas chromatography of transformer oil is analyzed without disintegrating transformer,the insulation performance of the transformer is diagnosed,and the insulation condition maintenance strategy can be further determined if other test methods are used.Therefore,it is of great academic significance to study the on-line monitoring system of dissolved gas in transformer oil.In this paper,firstly,the on-line monitoring method of dissolved gas in transformer oil is studied,secondly,the monitoring mechanism and existing problems of traditional oil chromatography on-line monitoring method are compared and analyzed,finally,a set of online monitoring system is designed,which can accurately monitor the dissolved gas in transformer oil.The simulation show that the on-line monitoring system can reflect the state of transformer insulation in real time.An in-depth study of data filtering algorithm and insulation fault diagnosis method is carried out for the chromatographic data of online monitoring,a white noise deconvolution algorithm for chromatographic peak recognition based on Kalman filter is proposed,and the fault diagnosis method based on wavelet neural network is also proposed.data processing model and fault diagnosis model are established and simulated.It can be seen from the experimental results that the white noise deconvolution algorithm based on kalman filter has higher accuracy and reliability.The fault diagnosis algorithm based on wavelet neural network is superior to traditional diagnosis method,it can be used to judge the deterioration of transformer insulation accurately and quantitatively,the equipment maintenance and maintenance is supported,it has certain practical value and application prospect.
Keywords/Search Tags:transformer, Dissolved gas in oil, On-line monitoring, Kalman filtering algorithm, Wavelet neural network
PDF Full Text Request
Related items