A Successive Geometric Segmentation Approach Applied to Double-Circuit Transmission Lines Fault Classification and Location

Abstract

This paper presents a new approach to address both fault classification and location in double-circuit power transmission lines. Fault diagnosis is achieved by using an algorithm based on the successive geometric segmentation approach. The proposed technique is able to generate both the topology and the weighting of neural networks. The input parameters are the magnitudes of phase voltages and currents measured in only one bus of a double-end fed transmission line. In order to validate the methodology, a comprehensive dataset of cross-country faults was simulated using a mathematical model. The results indicate high accuracy rate to fault diagnosis in double-circuit transmission lines compared to other ANN-based approaches found in the literature.