Design Of Online Monitoring And Fault Intelligent Diagnosis System For Transformer

To develope the smart grid,smart transformer will be applied in power system as a new generation of transformer.In order to guarantee the safe and economic operation,it's important to research on intelligent digital online monitoring and multiple information fusion diagnosis technology.It can be used for mastering real-time data,detecting latent fault and making maintenance strategy.Considering the variable transformer fault types,it is notreliable to adopta single on-line monitoring system based on the partial fault characteristics.In this paper,,intelligent transformer frame has been consteucted after a detailed study of the oil-immersed transformer.It's proposed that the intelligent integrated transformer on-line monitoring system is composed of five subsystems,such as the gas monitoring dissolved in transformer oil,partial discharge monitoring,winding temperature monitoring.The dissolved gas analysis in oil can only judge transformer fault qualitatively,it can not realize the fault location.Considering the limitations of traditional singal fault diagnosis method based on the dissolved gas in transformer oil,a method is proposed in this paper which is giving priority to dissolved gas analysis in oil,meanwhile combined with other diagnosis comprehensive fault diagnosis.In order to improve the diagnostic performance,it adopts the improved BP neural network to establish a fault diagnosis model in this paper.When establishing the fault diagnosis model,a way is put forward that optimizing the BP neural network by the differential evolution algorithm.It can improve convergence speed and precision immensely,meanwhile avoid falling into local extreme value.Finally,the on-line monitoring softwaresystemhas been designed.This system adopts Intouch as configuration software.It realizes data exchange with Oracle and Matlab through the ODBC and OPC technology.The system complete the integration of two sub-system,which are system status monitoring and fault diagnosis.The monitor canobtain the status of the transformer conveniently through this system.