Intelligent Methods for Fault Detection and Location in Transmission Lines
Fault detection and location in Transmission Lines (TLs) represent critical challenges in modern power systems, where even brief outages can cause significant economic losses and compromise grid reliability. Traditional protection schemes often struggle with high-impedance faults, varying fault locations, and complex operating conditions, demanding innovative real-time solutions. This work presents two advanced methods for fault analysis in TLs: one dedicated to fault detection and another to fault location, addressing these challenges through the extraction of Current Transient Deviation (CTD) curves obtained in real-time. These CTD curves enable the establishment of suitable threshold values for fault detection and serve as inputs to a Long Short-Term Memory (LSTM) neural network for precise fault distance estimation. The proposed solutions were tested using current signals from a 230 kV, 100 km TL and validated on the IEEE 9-bus system through RTDS real-time simulation. Extensive testing across varied fault resistances, incidence angles, and locations demonstrates robust performance regardless of fault characteristics. A comprehensive sensitivity analysis further validates the methods’ reliability under diverse operating conditions.