Laplace Deconvolution for Improvement of Signal Processing in Space Charge Measurement Applied to Full-scale Cables

This paper deals with signal processing in space charge measurement for full-scale High Voltage Direct Current (HVDC) cables using the pulsed electroacoustic (PEA) method. In cable applications, multiple reflections caused by acoustic mismatch between conductor and insulation system or from conductor bundles can lead to complex waveforms and long-term signal responses in the tail of the signal, which can be distorted after signal processing. Conventional Fourier deconvolution, which is commonly used to reconstruct the space charge distribution, suffers from baseline instability when the signal is cut off within a limited time window, making it unreliable in full-scale cable applications. To address this issue, we propose a numerical Laplace deconvolution method that effectively suppresses artifacts while preserving main signal components, thus improving signal reconstruction. The approach exploits the ability of the numerical Laplace transform to suppress unwanted reflections and improve baseline stability. To validate the proposed technique, experiments are conducted on a full-scale cross-linked polyethylene (XLPE) cable with 20 mm thick insulation. The reference signal without space charge effect is obtained by subtracting the measurements at different voltage levels, and the Laplace deconvolution is applied to calculate the charge density, electric field, and potential distributions at 50 kV bias voltage. The results show that the proposed Laplace deconvolution method significantly reduces the baseline variation and improves the measurement accuracy compared to the conventional Fourier deconvolution. This study demonstrates that Laplace deconvolution is a viable and effective alternative to improve the space charge measurements in practical HVDC cable systems.