Distribution High Impedance Fault Detection Using the Fault Signal Reconstruction Method

The high impedance fault has always been difficult to be detected for distribution grids due to the uncertainty of the weak fault signature. The fault signal reconstruction method is proposed and applied to detect the high impedance fault in the distribution grid. The key innovative feature of the proposed method is essential to the fault signal reconstruction with less prior knowledge and eliminate the effect of the scale decomposition algorithm on signal scale-level feature analysis to some extent. The fault signal is decomposed into multiple scale signals by the fast ensemble empirical mode decomposition. The feature extraction algorithm is proposed to extract scales with the essential fault features, and determine the coefficient of the selected scale signal. The signal reconstruction algorithm is proposed to reconstruct the coefficient of each scale signal for reconstructing signals detected for the sake of highlighting the fault characteristics within the raw fault signals. The signal reconstructed is detected by the principal component analysis. The feasibility and reliability of the method and the accuracy of the high impedance fault detection are verified by simulation results from PSCAD and Matlab with IEEE 34-node test system.