An Enhanced Feature Extraction Method for Power Cable FDR Test Results Based on First-Order Derivatives

Nowadays, frequency-domain reflection (FDR) cascaded with inverse Fourier transform (IFT) is widely used to detect the position of a power cable defect and diagnosis the type of the defect. However, due to the lack of the FDR test result in low frequency band width caused by the hard ware, it is hard to directly distinguish the type of the defect with the IFT result. Thus, an enhanced feature extraction method for power cable FDR test results based on first-order derivatives (FODs) is proposed to enlarge the difference of the IFT result between the general part and degraded part of the cable, and make the difference between the results characterized by different types of defects more obvious. Both simulated result and experiment result indicate that the proposed method can valuably enhance the features of cable defects in FDR test result, and reduce the error of defect type judgment caused by the lack of the FDR data within low-frequency band width and some other environment noise.