“A COMPARATIVE IMPLEMENTATION OF WAVELET TRANSFORM ALGORITHMS FOR TRANSIENT FAULT ANALYSIS”

Abstract

         Transient faults pose significant challenges in modern power systems and electrical networks due to their short duration, unpredictable nature, and potential to escalate into permanent failures if not detected promptly. Traditional signal processing techniques often struggle to accurately capture the time–frequency characteristics of these rapidly varying disturbances. This paper presents an efficient implementation of the Wavelet Transform (WT) for the precise detection, localization, and characterization of transient faults. By leveraging the multi-resolution capability of wavelets, the proposed approach decomposes the fault signal into different frequency bands, enabling detailed analysis of its dynamic behaviour.

In this paper, various mother wavelets were evaluated to determine the most suitable basis function for transient fault analysis, with Daubechies wavelets demonstrating superior performance due to their compact support and effective transient detection properties. The implemented WT algorithm processes sampled current and voltage signals to extract high-frequency components associated with fault inception. The detection method identifies sudden changes in wavelet coefficients, which serve as reliable indicators of transient disturbances.