A data-driven impedance estimation and matching method for high impedance fault detection and location of distribution networks

High impedance fault (HIF) is the most difficult fault type to recognize in power systems because their fault currents are small and difficult to distinguish from normal load fluctuations. Currently, most HIF identification methods are based on transient measurement data, and their dependence on high-frequency measurements, communication and data processing capabilities increases the cost and limits the application of these methods. This paper proposes a novel, data-driven two-stage HIF identification method based on line impedance estimation, which matches the estimated results with fault location information. The approach treats each line as a target, allowing for precise HIF identification between specific measurement nodes rather than at percentage locations along the full length of the line. Experimental results conducted on the IEEE 33 bus system show that the proposed method can pinpoint the occurrence of HIFs with 99.85% accuracy in the absence of noise and an accuracy of 91.72% with a signal-to-noise ratio of 60 dB for load identification results, demonstrating its effectiveness. • Detecting HIFs in distribution networks without high-frequency measurement. • Location of HIFs accurate to the line is achieved based on impedance matching. • Visual analysis shows varied HIFs occupy specific zones in impedance feature space.