Research On Fault Online Detection Technology Of Cable Based On Electromagnetic Coupling Method

As the upgrading and transformation of the urban transmission network, there is a growing demand for power cables. In order to ensure the safe operation of power grid and reduce the losses caused by failure due to cable insulation damage, the research on the online detection technology of cable insulation fault has important practical significance.Several existing and mainstream methods of the online detection are summarized and analyzed in this article, the detection technology of the power cable’s partial discharge is studied using the electromagnetic coupling method. First, the mathematical model of partial discharge signals is established, a simulation study on the transmission characteristics of partial discharge signals in XLPE cable is done using ATP-EMTP software, the rules on the propagation and attenuation of partial discharge signals are obtained.On this basis, a current sensor based on Rogowski coil used to capture the partial discharge pulse is designed. The electromagnetic parameters and the amplitude-frequency characteristics are deduced, and the amplitude-frequency characteristics are simulated in MATLAB. According to technology indicators, the coil parameters are calculated, and several coils are winded and verified by experiment.Partial discharge signal has the characteristics of wide frequency band and very small amplitude, and there is all kinds of disturbances that exist in the actual testing, and based on that, an amplifier and filter conditioning circuit of partial discharge signals is designed using operational amplifier with high gain-bandwidth product. By software simulation and experimental debugging, the conditioning effect to the partial discharge signals of the circuit is realized.As the performance limitations of the conditioning circuit and the electromagnetic interference at the detection site, filtering with FIR band-pass filter is studied, and using FFT to remove the periodic narrowband interference as well as using wavelet analysis to remove the white noise is simulated in MATLAB, then the denoise effect of the two methods is quantitatively analyzed in smoothness and similarity.