Neural Network Approach For Condition Monitoring of Traction Transformer oil

Abstract

As the development of high speed railway, it call for higher reliability of traction power supply system and traction transformer. However compared with pow transformer, on which the impulse load caused by sort circuit is continually applied, is ever worse. This bring 'Cumulative effect' which will give rise to different kind of transformer faults including hea discharged faults.. Hydro carbon gases and water are created in this course. By implementing correct operational and maintenance strategies the insulation aging/degradation process can be controlled and the traction transformer life can be extended effectively .A new approach has been presented in this paper based on  an index which could be computed online to monitor the status of traction transformer.