Research on Algorithm for Detecting Weak Submarine Cable Magnetic Signal Based on SAAFSA-SR Combined with CSASVD

Aiming at the problem that the magnetic signal is easily submerged in the noise due to magnetic attenuation in submarine cable detection, this paper proposes an algorithm for detecting weak submarine cable magnetic signal (WSCMS) based on stochastic resonance with self-adaptive artificial fish school algorithm (SAAFSA-SR) combined with cascaded self-adaptive singular value decomposition (CSASVD). SAAFSA-SR optimizes the strategies for the update of artificial fish school (AFS) parameters, and the strategies for artificial fish (AF) behaviors on the basis of the stochastic resonance with artificial fish school algorithm (AFSA-SR); and uses CSASVD to post-process the output of SAAFSA-SR, which includes the singular value decomposition (SVD) based on cascade principal component extraction and the adaptive separated order SVD based on differential spectrum (DS) analysis. After experimental verification, the SAAFSA-SR combined with CSASVD algorithm proposed in this paper can achieve significant enhancement of the local signal-to-noise ratio (LSNR) and global signal-to-noise ratio (GSNR) of target WSCMS.