Discrimination between inrush current and fault current by using wavelet transform

Abstract: Reliable discrimination between inrush currents and internal fault currents is crucial for effective protection of transformers in power systems. Traditional methods based on harmonic content analysis often struggle due to limitations in capturing both time and frequency domain characteristics of the current signals. This paper proposes a novel approach for inrush current and fault current discrimination in transformers using wavelet transform (WT).

Wavelet transform offers a significant advantage over conventional techniques due to its ability to analyse signals simultaneously in the time and frequency domains. This allows for the extraction of features that effectively differentiate between the transient nature of inrush currents and the sustained nature of fault currents.

The proposed method involves decomposing the transformer current signal using DWT (Discrete Wavelet Transform) and extracting relevant features from the decomposed coefficients. These features can then be used to design a decision rule or train a classifier to accurately discriminate between inrush and fault events.

Keywords: Internal Fault Current, Magnetizing Inrush current, Standard Deviation, Variance, Teager Energy Operator,Signal decomposition,Power transformer.