Método para detecção de curto-circuito fase-terra em redes secundárias de distribuição

Abstract

The detection of single-phase-to-ground faults in distribution systems represents a significant challenge, mainly due to the fault current, which is usually lower than in phase-to-phase and three-phase faults. Additionally, variations in system currents and voltages are less noticeable, making it more difficult to distinguish between an actual fault and normal operating conditions. This work presents a mathematical methodology developed using a simplified system to detect single-phase-to-ground faults, as well as training a machine learning model for detection. Although the use of machine learning is not the main focus of this work, it still demonstrates its importance and relevance in power systems. This study details the primary causes of faults in distribution lines, their impact on the power grid, and methods to locate and detect potential system failures. It also presents studies that emphasize the importance of an effective protection system and the use of artificial intelligence in power systems. Finally, the proposed methodology is explained, particularly the use of the inequality. The field of fault detection through mathematical equations remains underexplored. However, this work lays the groundwork for future research in the detection of single-phase faults.