Distribution Network Fault Location Method Based on Wide Scale Time Window Difference Operator

Abstract

This study aims to address the challenge of fault location in distribution networks with closed-loop structures. We propose a comprehensive method that focuses on two aspects: traveling wave head calibration and fault location. By analyzing the propagation properties of hybrid lines traveling waves, we demonstrate the feasibility of optimizing wave speed using time difference measurements. To solve the time calibration problem of weak fault signals, we utilize the signal cumulative deviation curve to select the time window of fault mutation, and then develop a novel wide scale time window difference operator for accurate calibration. Finally, we use a positioning method based on dynamic virtual faults to optimize wave speed and accurately locate faults in both overhead and cable lines. We build an improved IEEE 118 node network using power system simulation software (PSCAD) for simulation verification, which shows that our proposed method can accurately calibrate the initial wavefront time of fault traveling waves and precisely locate the fault positions.