ABSTRACT Diagnosing high‐impedance ground faults (HIGFs) in distribution networks is extremely challenging because high transition resistance significantly reduces electrical signal strength and unpredictable initial fault phase angles coupled with asymmetric voltage disturbances often lead to misclassification. This paper proposes a novel method for identifying single‐phase HIGFs based on spatiotemporal feature imaging of zero‐sequence voltage. By analysing and characterising the transient characteristics and phase‐space trajectories of zero‐sequence voltage in the complex frequency domain under both HIGFs and asymmetric disturbance conditions, a fault identification scheme is developed that integrates high‐ and low‐frequency information into spatiotemporal feature images. Leveraging a target recognition algorithm, the proposed method utilises low‐frequency image features for direct HIGF identification and high‐frequency image features to distinguish asymmetric disturbances from HIGFs in ambiguous cases, enabling automatic fault classification with enhanced accuracy. Experimental results demonstrate that the proposed method achieves superior performance, with classification accuracies exceeding 97% for HIGFs and 96% for asymmetric disturbances. Notably, it significantly outperforms conventional methods in distinguishing between asymmetric disturbances with specific initial fault phase angles and high‐impedance ground faults.
Li et al. (Sat,) studied this question.