Transformer Power Fault Diagnosis System Design Based On The HMM Method

Once failures for large-scale transformer power occur, which will result in catastrophic economic losses and social impact. Therefore, it is necessary to design and apply a state monitoring and fault diagnosis system for large-scale transformer in order to improve the reliability and accuracy for power transformer during its running, which will benefit increasing the power enterprise economic performance, promoting economic and social development. This paper aims at large-scale power transformer, introduces Hidden Markov Models theory into power transformer fault diagnosis field, and a fault diagnosis method using the HMM is put forward. Some issues and their settling methods about the HMM applying into power transformer brings are further analysed. The fault diagnosis principle bases on the HMM is discussed in detail. The power transformer faults are classified and each of their characteristic variables is determined, accordingly, the fault diagnosis model library for power transformer is researched. Finally, the fault diagnosis system for large-scale power transformer is designed.