Grid-Moving Equivalent Source Method in a Probability Framework for the Transformer Discharge Fault Localization

Abstract

The transformer discharge fault localization is very important for the power supply reliability and quality. The audible sound radiated by the transformer can reflect the extensive operation information. Thus, the discharge fault can be localized by high spatial resolution acoustic imaging techniques. In this paper, the grid-moving equivalent source method (GMESM) is proposed to localize the discharge fault on a transformer, which enables the equivalent source grids to approach the real position of the discharge fault adaptively. A probability framework is established in GMESM. The noise power can be firstly estimated by the Bayes rule to avoid an additional measurement of the background noise. Then, the expectation maximization algorithm is used to optimize the proposed model. Two experiment cases are used to verify the feasibility of the proposed method. It turns out that the proposed GMESM can localize the discharge fault on the transformer with high spatial resolution and accuracy.

Keywords: Grid-moving equivalent source method, Transformer discharge fault localization, Expectation maximization algorithm, Noise power estimation