Online Detection and Location Scheme for Aged Cable Segment in MMC-HVDC Systems Based on Admittance Spectrum and Compressed Sensing

The rapid development of offshore wind power has led to widespread uses of high-voltage direct current (HVdc) transmission systems with long-distance dc cables. The detection and location of aged cable segment is critical for ensuring overall system reliability. As current researches mainly focus on offline location methods, a novel online detection and location scheme for modular multilevel converter HVdc systems based on admittance spectrum and compressed sensing is proposed in this article. The relationship between admittance spectrum and aged segment location is mathematically modeled and impacts of converters are revealed. Then, for online extraction of spectra, an injection scheme is proposed, requiring no additional equipment. Finally, the idea of compressed sensing is used to cope with nonuniformly spaced spectra obtained online. Extensive simulations in PSCAD/EMTDC are conducted and average location errors are under 2%. Compared with existing approaches, the novelty of the proposed scheme lies in online implementation and high location precision.