Edge Computing-Based Fault Location in Distribution Networks by Using Asynchronous Transient Amplitudes at Limited Nodes

The feeders in distribution networks can be generally classified into two types: homogenous and hybrid feeders. Hybrid feeders consist of non-homogeneous line segments. In this paper, a fault location scheme is proposed for distribution grids with both homogenous and hybrid feeders using the edge computing. The standard transient amplitude ratio arrays of the hybrid feeders are first constructed and stored at the edge nodes. Then, an accurate fault location model is trained by the radial basis function neural network and stored in the substation server. To achieve this goal, the transients are first sampled at the measurement nodes and transmitted to the edge nodes nearby. The amplitudes are then extracted to determine the faulty branch. For a hybrid feeder, the actual amplitude ratio array is compared with the standard one to identify the faulty line segment. Finally, the amplitude ratio of the two transient components with different frequencies at one terminal of the faulty section is computed and transmitted to the server. Based on the model of a distribution network built by PSCAD/EMTDC, extensive fault simulations are studied. The results demonstrate the effectiveness of the scheme.