A Data-Driven-Based High Impedance Fault Location Method Considering Traveling Waves in Branched Distribution Networks

Distribution utilities responsible for supplying electricity face challenges in maintaining continuity of power supply and serving end consumers. Locating faults in the distribution network is a major challenge since faults can cause long-duration disruptions in power supply. Therefore, effective fault localization techniques, particularly for High Impedance Faults (HIFs), have become an area of focus. These faults are not detected by conventional protection equipment as they have currents with magnitudes similar to loads in the distribution network. Despite the considerable research efforts devoted to this issue, there is still no universal solution to locate such faults. Thus, this paper proposes a methodology that can identify the branch of HIFs in the distribution network and determine its location precisely. The proposed technique uses traveling wave method, Discrete Wavelet Transform (DWT), and Artificial Neural Network (ANN) as machine learning method. The proposed method has undergone numerous tests considering various inception angles, load variations, and different networks to prove its robustness and effectiveness. The results show that the proposed method is promising, with a high accuracy for determining faulty section and a low error ratio for fault distance calculations.