Local defect location method for long cables based on Improved MUSIC-Pseudospectrum

Reflectometry (FDR) suffers from resolution limitations due to random defects distribution and severe signal attenuation that hampers the detection of weak local defects, this paper proposes a localization method for partial defects based on the real‑valued Multiple Signal Classification

(MUSIC) transform. Firstly, a reflection and refraction model of cable defects is established based on transmission line theory, and the quantitative relationship between defect positions and the reflection coefficient spectrum is derived. Secondly, an eigen decomposition method with real-valued covariance matrix is proposed, then the signal subspace dimension is adaptively estimated using the relative energy threshold of eigenvalues, and a pseudo-spectrum function is constructed based on the orthogonality of the noise subspace, enabling effective localization of multiple local weak defects in long cables. Finally, the proposed method is compared with traditional methods in a 1000m cable simulation.

Results show that, with a maximum test frequency of 20 MHz, the proposed method accurately locates single and multiple mixed defects with a maximum localization error of 0.1%. Compared with the traditional Fourier transform method, the proposed method achieves super-resolution localization under low bandwidth, overcoming the constraints of high-frequency signal attenuation on detection accuracy, and provides an effective means for early warning of weak defects in long-distance power cables.