Accurate fault location methods can greatly reduce the power recovery time, which has an important effect on the power supply reliability. Fault location methods have been extensively studied; however, location for high impedance faults and faults with near-0° fault angles or low signal-to-noise ratio (SNR) transient waves remain challenging. In this paper, a fault location method based on edge detection is proposed. This approach extracts fault features using the Canny operator and determines the fault inception time using the Otsu threshold. The algorithm is described using the noise-free and noisy fault waveforms as examples. The parameters in the algorithm can be adaptively adjusted according to the frequency and SNR of the fault waveforms. A 500 kV AC transmission line was built to compare the performance of the proposed method with that of the fault location method based on the wavelet modulus maximum under different conditions of fault resistance, fault position and fault angle. The results show that the proposed method not only has a better location accuracy under typical fault conditions, but is also more robust for faults with low SNR traveling waves such as those with high fault resistances and near-0° fault angles.

准确的故障定位方法可以大大缩短电源恢复时间，对供电可靠性有重要影响。故障定位方法已被广泛研究；然而，定位高阻抗故障和故障角接近 0° 或低信噪比 (SNR) 瞬态波的故障仍然具有挑战性。本文提出了一种基于边缘检测的故障定位方法。该方法使用 Canny 算子提取故障特征，并使用 Otsu 阈值确定故障起始时间。该算法以无噪声和有噪声故障波形为例进行描述。算法中的参数可以根据故障波形的频率和信噪​​比进行自适应调整。搭建了一条500 kV交流输电线路，在不同的故障电阻、故障位置和故障角度条件下，比较了该方法与基于小波模最大值的故障定位方法的性能。结果表明，所提出的方法不仅在典型故障条件下具有更好的定位精度，而且对于具有低SNR行波的故障（例如具有高故障电阻和接近0°故障角的故障）也具有更强的鲁棒性。