Principle and Verification of an Improved Algorithm for Cable Fault Location Based on Complex Reflection Coefficient Spectrum

Yuxiao Hu,Lu Chen,Ying Liu,Yang Xu
DOI: https://doi.org/10.1109/tdei.2022.3224700
IF: 2.509
2023-01-01
IEEE Transactions on Dielectrics and Electrical Insulation
Abstract:Diagnostic methods for fault location in cable insulation have become increasingly vital with the development of long-distance cables. This article presents an improved method for fault location based on the complex reflection coefficient spectrum (CRCS). First, a mathematical model for a long high-voltage cable line with consideration of a distribution-parameter circuit of cable joints is established in the frequency domain, and the CRCS of the cable with joints is analyzed. Next, based on the principle of frequency domain reflectometry (FDR), an algorithm in time-frequency conversion which uses the CRCS as the analysis spectrum is proposed. And the positioning sensitivity of the method is improved by implementing a data process adopting a Chebyshev window function which is controllable and reduces the sidelobe effect effectively. Furthermore, the improved method is evaluated by performing the simulations and two field verification tests. In a field test for a 5.1-km-long 110 kV cable, the multiple cable joints could be located effectively, and the relative error is between 0.14% and 1.49%. Then, the method was conducted for the positioning of a high-resistance fault and multiple joints on a 2.7-km-long 110 kV cable and the relative error is 0.37%. It is worth noting that the cross-bonding links of both cable tests were artificially replaced with temporary links straight across each joint. For the cross-bonded cable system, the test result shows that an obvious reflection is found at the cross-bonding point, which can reduce the reliability and accuracy of the present method.
What problem does this paper attempt to address?