44. Fault Diagnosis for Power Transformer Mainly Based on Power-On State Variables

Nowadays, the synthetic diagnostic methods using both power-on and power-off state variables are widely studied and applied. The usage of multiple state variables can increase the accuracy of diagnosis, but it usually brings many problems into practical application at the same time. The main problem is that the more the power-off state variables, the longer the time without power, and the worse the reliability of transformers. Therefore, this chapter presents a diagnostic method mainly based on power-on state variables. It consists of primary diagnosis and precise diagnosis. Primary diagnosis uses only power-on state variables to acquire a part of the fault information. Then, in order to get the exact fault types and more position information, based on the results of primary diagnosis we select and use some useful and necessary power-off state variables in the precise diagnosis stage. And during this process, the rough set theory is used to predigest knowledge and reduce the complexity of diagnosis. Compared with other diagnosis methods, this method shortens the outage time and improves the efficiency of diagnosis. In the end, the diagnosis is proved to be practical and effective by a fault case.