Neural-network-based fault location estimator for transmission line protection

Article type: Research Article

Authors: Altuve, Héctor J. | Chacón, Oscar L. | Vázquez, Ernesto | Posadas, Daniel | Sánchez, Edgar N.^;

Affiliations: Facultad de Ingeniería Mecánica y Eléctrica, Doctorado en Ingeniería Eléctrica, Universidad Autónoma de Nuevo León, P.O. Box 89-F, Ciudad Universitaria, San Nicolás de los Garza, Nuevo León, C.P. 66450, MEXICO. E-mail: evazquez@gama.fime.uanl.mx

Note: [] Present address: Cinvestav, Unidad Guadalajara, Apartado Postal 31–438, Plaza La Luna, Guadalajara, Jalisco, C.P. 45091, Mexico

Abstract: A classical task in the protection of transmission lines against short circuits is the estimation of the electrical distance to the fault and its comparison against a given threshold to determine whether the line is faulted or not. This paper presents a novel neural network approach to this problem, as a step towards the design of a neural protective relay. Two different alternatives are proposed and evaluated in the paper, which resemble the amplitude and phase signal comparison principles currently used in analog protective relays. Obtained results confirm that neural network technology can successfully be used to estimate the fault location for transmission line protection.

Journal: Journal of Intelligent and Fuzzy Systems, vol. 7, no. 2, pp. 159-171, 1999