Power System Fault Detection and Classification: Case Study with Large-Scale Grid Data

Rapid and accurate fault classification is crucial for ensuring system stability and minimizing downtime. This paper focuses on real-world cases from disturbance records. A novel algorithm is developed for the detection of abrubt changes. With that algorithm, 3-phase, phase to phase and single phase faults are identified. Further characteristics, such as maximum short circuit currents and inception angle are collected. The proposed algorithm is tested and benchmarked against existing methods based on symmetrical component analysis, demonstrating superior performance in terms of speed and classification accuracy.