Partial Discharge Localization Based on Direct Data-Reusing TDE Algorithm With Generalized Cauchy Loss

The presence of impulsive noise can significantly compromise the localization accuracy of partial discharge (PD). To address this problem, the direct data-reusing time-delay estimation (TDE) algorithm with the generalized Cauchy loss (GCL) function (DDR-TDE-GCL) is proposed. Leveraging the robustness of the GCL function, the proposed algorithm demonstrates a superior capability to combat impulsive interference. In addition, the direct TDE method is integrated into the data-reusing strategy. The integration enables the DDR-TDE-GCL algorithm to achieve high localization accuracy in cable systems while remaining noniterative. Moreover, the convergence behavior of the DDR-TDE-GCL algorithm is analyzed, proving that the estimation for the attenuation factor is unbiased and consistent. Simulations and experiments demonstrate that the DDR-TDE-GCL algorithm has an improved localization accuracy compared to existing algorithms for PD signals, achieving an average localization accuracy of 99.59%.