Fault Induced Transient Detection in Power Systems: An Adaptive Approach Considering Noisy Environment

This paper proposes a new approach for transients detection, specifically for faults-induced transients of common and high impedance faults. The approach is based on the DQtransformation with an adaptive filtering technique in order to generate a detection signal. Transients detection is made by means of a mathematical algorithm which is self-adaptive to the noise level of the detection signal. The proposed algorithm is based on the analysis of two dynamic thresholds. The thresholds definition is made with an accumulating confidence decision formulation in order to verify true detections. The proposed approach is evaluated using simulated cases of high impedance faults on a model of a real distribution system. Comparisons with the last published method based on DQ-transformation show that the proposed approach is more robust to noise.