Application technique for model‐based approach to estimate fault location

Abstract

Impedance-based algorithms commonly used for determining the fault location in transmission lines are prone to several sources of error and are specific to the line and system configuration. Furthermore, these algorithms do not utilise available valuable information about the power system surrounding the faulted line. These issues can be overcome using a model-based fault location (MBFL) approach. It uses a circuit model to simulate possible fault scenarios and compares the simulated fault currents with the measured currents recorded by the relay to identify the fault location. However, there are several difficulties and limitations while applying MBFL. There is a loss in accuracy and precision based on the number of simulated scenarios and a requirement to store voluminous simulation results. Hence, this study presents a novel application technique for implementing model-based approach efficiently to estimate the fault location and fault resistance using artificial neural networks-based approach. A key highlight of the proposed approach is the ability to identify the location of a fault present on neighbouring lines using the measured through fault current. The study also presents representative scenarios to demonstrate the capability and potential of the proposed approach.

Accuracy Estimation for Operating Characteristics Neuromodeling of the Overcurrent Protection in a Three Phase Mains

“…Для этих целей могут использоваться методы машинного обучения, в частности аппарат искусственных нейронных сетей (ИНС) [7][8][9][11][12][13], как одна из основ искусственного интеллекта [14,15]. Эти инструменты могут использоваться для моделирования [24], обработки сигналов [17,20,23], анализа информационного обеспечения и информационного обмена в электросетевом оборудовании [16]. В полной мере это относится к устройствам релейной защиты и автоматики [26], задачам определения места повреждения [1,4,22].…”