Location Method for Asymmetrical Latent Cable Faults in Low-Resistance Systems Based on Multidimensional Information

The incipient cable fault in active low-resistance-grounded distribution networks is a typical asymmetrical fault and is difficult to locate because the fault current is weak, short-lasting, and easily affected by distributed generation (DG). To address this typical asymmetrical problem, this paper proposes a fault section location method based on multidimensional information correlation analysis. First, an equivalent incipient fault model is established by combining the Kizilcay arc model with an insulation-defect resistance, so that the intermittent arc behavior and the conductive path of degraded insulation can be represented simultaneously. Then, the generalized S-transform is used to extract three features from the transient zero-sequence current, namely the transient current energy index, group phase-angle polarity, and waveform similarity. On this basis, a multidimensional feature vector and a comprehensive similarity coefficient are constructed to identify the fault section, and an auxiliary downstream energy comparison rule is introduced to distinguish the actual fault section from DG-connected pseudo-fault sections. The method is verified in MATLAB/Simulink R2025a under different fault locations, DG access conditions, penetration levels, noise levels, and key parameter variations. The simulation results under the tested conditions indicate that the proposed method can effectively identify asymmetrical incipient cable fault sections in active low-resistance-grounded distribution networks.