It is difficult to judge the polarity of cable impedance discontinuities due to the limitation of the hardware lower testing frequency when using frequency domain reflection (FDR) technology for defect detection of long cable.  In this paper, a new time-frequency pulse conversion algorithm is proposed based on FDR for cable impedance discontinuities to realize defect location and polarity discrimination.Firstly, the cable reflectance spectrum is obtained, and then the impedance discontinuities of the cable are located by using the second-order Nuttall self-convolution window and fast Fourier calculation. Secondly, the time-frequency pulse conversion algorithm of cable impedance discontinuities is proposed. The method makes use of the good frequency adjustment ability of Gaussian narrowband envelope signal to adjust its parameters so that the frequency domain center of pulse signal is located in the test band of sweep signal, and the optimal pulse width is set to achieve the minimum range resolution at the same time. According to the time-frequency characteristics of Gaussian envelope narrowband signal propagating in the cable, the main peak of the time-frequency pulse is determined, and the polarity of the local defect of the cable can be judged according to the polarity of the main peak.Finally, the proposed algorithm is verified by the experiment of the 10 kV XLPE cable with a joint at 1500 m in the laboratory.The results show that the time-frequency pulse conversion algorithm can effectively locate the impedance discontinuities of the long cable and accurately identify the fault type.

Long cable; Impedance discontinuities; FDR; Polarity discrimination; Time frequency pulse conversion