State Evaluation of 500 kV Transformer in DC Bias Based on Information Fusion

In order to accurately evaluate the direct current (DC) bias level of the transformer, an evaluation method based on information fusion is proposed in this paper. Firstly, the features of vibration signals in the time, frequency and time-frequency domains under different severity of DC bias are collected and analyzed. Then, a multi-level evaluation model combining radial basis function (RBF) artificial neural network and improved analytic hierarchy process (AHP) is established. The RBF neural network takes the features as input and outputs the probability of the DC bias level initially. Then, the outputs of RBF are assigned certain weights via improved AHP, and the evaluation of DC bias of transformers can be realized. The proposed method is verified based on field data from 500 kV transformers of China Southern Power Grid. The results show the accuracy and feasibility of the evaluation method.