Research on fault diagnosis method of distribution transformer based on MFCC and HMM

The acoustic signal of the distribution transformer contains a much of information about the state of the transformer operation. Therefore, the fault diagnosis method of distribution transformer based on MFCC and HMM is proposed. The acoustic diagnosis process of transformer fault can be divided into three parts: (1) pre-acquisition and pre-processing of signal, (2) the MFCC and its first-order difference coefficient of the transformer acoustical signal are extracted as the characteristic quantities of the acoustic signals of distribution transformers, (3) the hidden Markov model (HMM) is used to classify and identify the acoustic signals. This method is verified by experiments and the influence of several common environmental disturbance on fault diagnosis is considered. The results show that the method can be applied to the fault diagnosis of the distribution transformer, and it has a certain anti-interference ability.