Novel Parameter Identification Method of Extended Debye Model for Long-Length Submarine Cables

Abstract

To address the challenge of acquiring detailed insulation status for long-length submarine cables, this article proposes a branch parameter identification method without manual thresholds, based on an extended Debye model (EDM) and the polarization depolarization current method. First, depolarization current is used to construct a Hankel matrix, and the singular values and incremental sequences of information entropy are acquired by step. Subsequently, the matrix pencil (MP) algorithm is employed to determine the branches with the highest amplitudes, which are successively removed from the original signal. The number of branches is confirmed by minimizing the variance of the information entropy increment sequences according to the characteristics of white noise. Simulation results demonstrate that the proposed method achieves high accuracy even at SNR =30 dB. Moreover, the proposed method remains accurate even when there are significant differences in branch amplitudes. Finally, the proposed method is applied to four in-service submarine cables, and the results are corroborated through comparisons with data obtained from the frequency domain reflection (FDR) method and temperature-measuring optical fiber; thus, validating the effectiveness of the proposed method.