Real-time Detection Approach: Applied To Potential Transformer-based Over-voltage Monitoring

摘要

Potential transformer(PT)-based over-voltage monitoring has the advantage of low cost and safety. However, signal distortion due to poor frequency characteristics makes overvoltage detection difficult. This paper introduces a software-based real-time overvoltage detection approach. First, the voltage signal data continuously acquired at high speed are taken as input, and the feature point sequence(FPS) is obtained by extracting local extremum points of the signal. Then, based on the FPS, characteristic parameters such as peak voltage (PV), maximum waveform gradient (GM), number of pulses (PN), and waveform symmetry coefficient (WSC) are calculated. Finally, these feature parameters are fused to obtain a normalized overvoltage detection coefficient (ODC). The ODC is compared to a set threshold to determine if an overvoltage has occurred. Multithreaded programming techniques are used for real-time detection of over-voltages. The monitoring system using this approach has been successfully applied.