Search

To improve the reliability of power systems, online cable condition monitoring must be implemented to replace traditional regular power outage maintenance. Existing online monitoring methods have certain limitations. In this article, a novel method based on low-frequency signal injection is proposed to monitor power cable insulation conditions online. For this method, a low-frequency signal is injected into the power system via the potential transformer (PT) open delta configuration. The cable conductor voltage and leakage current are detected. The interpolating windowed fast Fourier transform (FFT) algorithm is applied to calculate the dielectric loss angle. Then, the cable tangent delta ($\tan \delta $) can be deduced, and the cable condition can be assessed. A distribution network system is built based on the power system computer-aided design (PSCAD) software. The frequency and amplitude of the injected signal are simulated and discussed. An online monitoring platform is established in the laboratory. Thermal-aged cable samples and water tree-aged cable samples are prepared and measured. The experimental results confirm that the proposed method can effectively assess power cable insulation. [ABSTRACT FROM AUTHOR]

The onsite partial discharge (PD) detection of cable insulation is of great importance for power system maintenance. To preserve more basic features of denoised PD signal, in this paper, a PD denoising by a new class of affine projection algorithm (APA) is proposed based on the high-order error power (HOEP) criterion. We propose a new affine projection least mean absolute third algorithm (APLMATA), which exploits the HOEP criterion ($p=3$) and can achieve reliable performance under Gaussian interference. We further develop a modified APLMATA (MAPLMATA) to reduce the misadjustment with moderate computational burden. Analytical models are derived for MAPLMATA for Gaussian signals. To enhance the stability of the algorithm in generalized Gaussian noise scenarios, an affine project least mean $p$ th power algorithm (APLMPA) that includes the $p$ th power dependence on the error is presented. In particular, we apply the proposed APA-based algorithms to suppress the noise from PD signal in power cables. Simulation and experimental results demonstrate the effectiveness of the proposed algorithms for system identification and PD denoising. [ABSTRACT FROM AUTHOR]