Data-Driven Transmission Line Fault Location with Single-Ended Measurements and Knowledge-Aware Graph Neural Network

Transmission line fault location is one of the essential steps to ensure power supply reliability. Traditional model based methods and traveling wave based methods have limitations such as requirements of accurate line models/parameters or high sampling rates. On the other hand, most existing data-driven methods only utilize information within the raw data and fail to adopt prior physical knowledge. This paper proposes a data-driven fault location method based on knowledge-aware graph neural network (GNN) with single-ended measurements. Firstly, for different fault types, the graph structures are carefully designed to represent the inherent relationship among the measured voltage, measured current and the fault location, to incorporate prior physical knowledge. Afterwards, the GNN is adopted to achieve line fault location. The method only requires single-ended three phase voltage and current instantaneous measurements, with a relatively low sampling rate of 80 samples/cycle according to IEC61850-9-2 standard. Numerical experiments prove that the proposed GNN based fault location method has higher fault location accuracy compared to the existing multilayer perceptron (MLP) based fault location method.