Credibility-Based Fault Fusion Diagnosis Method for Power Transformer

Abstract

The fault causes and phenomena of power transformers are complex; different monitoring parameters can reflect its state from different dimensions. In view of the diversity of monitoring data and the characteristics that the acquisition time of different monitoring data is not completely simultaneous, the power transformer fault fusion diagnosis method is proposed based on credibility theory in this paper. First, the monitoring data, including dissolved gas analysis data, partial discharge pulse signal, ultrasonic signal and vibration signal, are diagnosed, respectively. Then the credibility method is used to fuse the single diagnosis results to obtain the comprehensive state of transformers. The case analysis shows that the proposed method can not only make full use of various monitoring data to enhance the accuracy of transformer fault diagnosis but also apply to the fault diagnosis of transformers with different monitoring data types. This provides a new idea for the fusion diagnosis of power transformers.