Sensitivity Analysis of Water Tree and Input Pulse Parameters for Time-Domain Reflectometry of Power Cables Using Taguchi Method

In this paper, quantitative sensitivity analyses of various parameters of water tree and input pulse are performed using Taguchi-based design of experiment method in order to verify the feasibility of time-domain reflectometry (TDR) method for the water tree detection in the power cable. Signal to noise ratio (SNR) and percent contribution (PC) of water tree and diagnostic pulse parameters are calculated, and the parameters are identified that have high influence on the reflected pulse. The simulation software (COMSOL Multiphysics) used for the analysis is verified by comparison with the experimental results of the cable with an artificial defect. Based on the sensitivity analysis, optimal diagnostic pulse parameters are proposed for the efficient detection of the water trees.

Automated summary

This paper conducts quantitative sensitivity analyses of various parameters of water tree and input pulse to verify the feasibility of the TDR method for water tree detection in power cables. Optimal diagnostic pulse parameters are proposed based on the sensitivity analysis for efficient detection of water trees. The simulation software used is verified by comparing with experimental results of cables with artificial defects.