Edge Computing Based Fault Sensing of the Distribution Cables Based on Time-Domain Analysis of Grounding Line Current Signals

Abstract

Due to the complex structures and fault conditions, accurate fault sensing of the underground distribution cables is a difficult task. The distribution automation system (DAS) and fault indicator system (FIS) are only applicable to the severe faults. In addition, the current DAS and FIS are all based on the centralized data processing system (CDS), increasing the communication pressure and data processing time of the main station. Considering the advantages of edge computing, this paper investigates using the grounding line currents to sense the fault in distribution cable networks. First, an edge computing based fault sensing system (ECS) is designed. The equivalent time-domain distribution cable models considering multi-conductor mutual couplings are then established in both normal and fault cases. The time-domain features of the cable grounding line currents prior to and after the fault are analyzed to create the fault detection criterion. Combining the network topology, locations of the measurement units (MUs), and direction factor of the ground line current, a space matrix based algorithm is created to achieve fault cable section location at edge. Finally, a distribution cable network is created by PSCAD/EMTDC. Various simulations are conducted to validate the effectiveness and robustness of the algorithm.