Advanced signal analysis for high-impedance fault detection in distribution systems: a dynamic Hilbert transform method

This paper presents a novel approach for detecting high-impedance faults (HIF) in distribution systems that uses the Hilbert transform. Our approach is based on determining the instantaneous frequency of signals and detecting deviations from a reference frequency. Our technique is very sensitive to fault fluctuations because it makes use of the Hilbert transform's ability to capture dynamic signal properties like phase and frequency alterations. This sensitivity enables the extraction of unique features that identify fault signals, providing critical insights into fault detection and location. Notably, our method is appropriate for the analysis of non-stationary signals, which are typical in power systems where signal attributes vary fast during fault conditions. Furthermore, our method resolves deviations by comparing them to a predefined range and displaying essential features such as basic frequency, RMS (Root Mean Square), Crest Factor, Minimum and Maximum Deviations, and Maximum Current Amplitude. These values offer unique insights into the present signal's qualities, which aids in defect detection and diagnostics, particularly in HIF settings. Our proposed technique detects high-impedance flaws by evaluating deviations from the nominal frequency, even in environments with weaker features and variable surface conditions. To improve our system's robustness and usefulness, we recommend performing additional study on adaptive thresholding algorithms and real-time implementation choices. Future research areas could involve investigating the integration of machine learning algorithms for automatic fault categorization and localization, which would enhance the capabilities of distribution system fault detection approaches.