The current time domain spectroscopy of dielectrics provides important information for the analysis of dielectric properties and mechanisms. However, there is always interference during the testing process, which seriously affects the analysis of the test results. Therefore, the effective filtering of current time domain spectroscopy is particularly necessary. L1 trend filtering can estimate the trend items exactly in a set of time series. It has been widely used in the fields of economics and sociology. Therefore, this paper attempts to apply L1 trend filtering to the current time domain spectroscopy. Firstly, polarization and depolarization currents are measured in the laboratory. Then the test results are filtered by L1 trend filtering and the filtering effects are compared with several common filtering algorithms, such as a sliding mean filter and Savitzky–Golay smoothing filter. Finally, the robustness and time complexity of L1 trend filtering are analyzed. The filtering results show that because the polarization currents vary in a wide range of the time domain (about 2–3 orders of magnitude), smooth and undistorted curves in the whole test time range can hardly be obtained through common filtering algorithms, while they can be obtained by L1 trend filtering. The results of robustness analysis and time complexity analysis show that L1 trend filtering can extract the trend items accurately in the time series under given different noise levels, and the execution time is also lower than 176.67 s when the number of tested points is no more than 20,000. Those results show that L1 trend filtering can be applied to the time domain current spectroscopy of dielectrics.

The insulation resistance (IR) test has been widely conducted by electricity companies to assess cable insulation health status due to its ease of applicability. However, there exist cases in which medium-voltage power cables pass the IR test but fail in service after re-energization. So far, historical IR data obtained via measurements have been recorded, but have never been systematically studied to improve the accuracy of detecting unhealthy cables. This paper proposes a data analytic approach using historical IR data to identify patterns for distinguishing between healthy and unhealthy cables and assess the cable insulation condition. The proposed approach first leverages a two-parameter Weibull analysis to link the failure probability of the cables to their ages. Such analysis sheds light on the classification of the cables with respect to their age and material. Next, a diminishing method (DM) is used to set the critical IR values and provide maximum detection of the unhealthy cables with minimum misclassification of the healthy cables as unhealthy. Finally, a self-organizing-map-based support vector machine (SOM-SVM) is used to classify the cables as healthy or unhealthy. The hybrid DM-SOM-SVM approach is applied to the historical IR data of 22kV and 6.6kV cross-linked polyethylene (XLPE) cables. Compared to current industrial IR criteria for insulation condition diagnosis, the proposed approach allows detecting 18.5x and 1.8x more unhealthy 22kV and 6.6kV XLPE distribution cables, respectively.

The calculation of an electric field distribution provides the basis for the structural design of the insulation, and an accurate characterization of conductivity as a function of temperature and electric field forms an important basis for the simulation of the electric field distribution in HVDC (high-voltage direct current) cables. However, the conductivity functions that describe the insulating materials used for HVDC cables in different studies are different, and very little has been reported regarding how to choose the most accurate function. In this work, the conductivity of insulating materials used for HVDC cables is characterized, and the effects of the conductivity characterization on the simulation of the electric field in HVDC cables are studied. First, eight common conductivity functions are compared qualitatively. Then, the conductivities of XLPE for different temperatures and electric fields are measured, and a data fitting technique is used to analyze the coincidence degree between different functions and the test results. Finally, the steady-state electric field distributions of HVDC cables for different temperature gradients are simulated in COMSOL Multiphysics. The results show that the sum of the square of the relative errors of the fitting when using the original functions is larger than that achieved when using the logarithmic form of the functions. The deviations in the electric field caused by taking the logarithm of different functions are smaller.

This article took 6.6-kV medium-voltage crosslinked polyethylene (XLPE) cables as the test object, conducting field tests of polarization and depolarization currents (PDCs), very-low-frequency (VLF) tan $\delta $ , and time-domain reflectometry (TDR). The correlation between PDC testing and VLF tan $\delta $ and TDR measurements was investigated to enhance the effectiveness of the PDC testing method. PDC test outcomes indicated that polarization currents sometimes exhibit increasing or non monotonic trends over time, referred to as abnormal trends herein. Through in-depth research combining TDR measurement results and relevant theoretical frameworks, it is confirmed that abnormal polarization currents are attributable to water ingress issues in cable joints. Water ingress into the joint induces nonlinear variations in insulation conductivity with the electric field, manifesting as abnormal trends in polarization currents at the joint-insulation interface. These kinds of abnormal polarization currents can serve as an effective diagnostic indicator for identifying cable joint issues. Moreover, VLF tan $\delta $ measurements alone sometimes fail to accurately diagnose water ingress issues in cable joints, necessitating the use of TDR for further diagnosis.

Effective health assessment and management of underground power cable system is essential for ensuring the cost-efficient operation of power grids. Existing practices usually monitor condition of the overall health of the power cable circuits but overlook its intricate structural complexity. This letter proposes a data-driven health assessment method considering both failure-related structural characteristics of each cable circuit and condition measurement data. Firstly, distinguishing failure indicators are identified by integrating in-depth component analysis with circuit condition measurements. Based on these features, a novel integrated health index is proposed, named as One-sample Maximum Mean Discrepancy (O-MMD), which evaluates the health condition of a circuit by quantifying the disparity between un-faulted assets and typical faulted groups in a high-dimensional feature space. This approach is able to assess whether new or modified circuits might exhibit similar fault characteristics as previously observed cases. Based on the O-MMD index, the probability of failure can be estimated for each cable circuit, and the maintenance plan can be prioritized. The proposed method is demonstrated on both seen and unseen real-world underground cable system data in Singapore.

Health assessment of underground cross-linked polyethylene (XLPE) power cables is crucial for maintaining power grid reliability. Existing methods are mostly based on cable condition data, without sufficient utilization of water trees formed in the cables, a very important health indicator. To make full use of heterogenous data of cable health information, this letter proposes a multimodal large-language model (LLM) method for cable health assessment, which simultaneously uses water tree images and condition monitoring data, including insulation resistance (IR), IR Ratio, very low frequency (VLF) tan ${\bm{\delta }}$, and age, as the input and generates a quantitative health index. For training data collection, lab experiments are conducted to capture water tree images of retired cables. Then, the images and cable condition data are combined to fine-tune a LLM through low-rank adaptation (LoRA), which is highly computationally efficient and requires much less memory. The proposed method is implemented with Gemma 3-4B and tested on real cable data in Singapore.