Mathematical-detection-algorithms for cable diagnostics and investigations with Line-Resonance-Analysis

For power cables several diagnostic methods are available. The main and typical diagnostic methods can be separated into global and local failure detection systems. An example for the global detection systems is the Dissipation-Factor-Measurement (tanδ). This technique is well established in the diagnostic and allows detecting a global failure, e.g. an electrical tree. However, this technique can't localize the failure position. Approved detection systems for local failures are the Time-Domain-Reflectometry (TDR) and the Partial-Discharge-Measurement (TE). These techniques need a failure (small insulation discontinuity or impedance skip) to detect the position. The Line-Resonance-Analysis (LIRA) is a possible solution for the detection of global and local failures. The system isn't well established and controversially discussed by experts. In this research study we investigate different algorithms applied to the LIRA in order to clarify the abilities of the LIRA. The LIRA technique is a modified type of the Frequency-Domain-Reflectometry (FDR). The main difference is that the LIRA uses the information of the resonance frequencies. The investigation of the LIRA is based on spectral analysis, which uses the Fourier-Analysis (DFT/FFT). Unfortunately this method is limited in the spacial resolution. For this reason and the fact that other investigation techniques for diagnostic use other possible detection algorithm, we want to examine more feasible mathematical functions. Potential algorithms for the spectral analysis are the Goertzel-Algorithm and the Chirp-Z-Transform (CZT). In this paper the research focus lies on the LIRA with different possible detection algorithm.