Anti-noise UWFBG hydrophone array enhanced DAS system using optimising SVMD based on matched interferometric

Abstract

To address the problem of the ultra-weak fiber Bragg grating (UWFBG) background noise in the high-performance interferometric fiber-optic hydrophone sensing system, this paper proposed a combination of RIME optimization algorithm and successive variational mode decomposition (RIME-SVMD) algorithm to mitigate the non-stationary noise signal with time-varying frequency components, which induced by light source and other non-ideal optical transmissions during 3 × 3 demodulation. The SVMD algorithm adopts an iterative approach to decompose the demodulation signal into several Intrinsic Mode Function (IMF) components. Althouth it is worth noting that SVMD algorithm entail a fewer parameter without selecting the number of modes compared to traditional VMD algorithms, it is improved by RIME algorithm to establish the optimization objective, whose phase permutation entropy (PPE) mean of the modal components get minimized to realize the self-adaptive selection of parameters. The proposed algorithm is applied to the adaptive decomposition of the underwater acoustic, resulting in multiple IMF components. The IMFs with the maximum correlation coefficient criterion are retained for signal reconstruction, which contain a significant amount of underwater acoustic characterisation information. The experimental and comparative results show the exceptional performance of the proposed algorithm in effectively removing system and environmental noise while preserving underwater acoustic information. Furthermore, the signal-to-noise ratio (SNR) of the acoustic signal is improved by 6.42 dB compared to VMD. Furthermore, through the information process of beamforming the array signals before and after noise suppression, we note that the approach is capable of achieving directional detection by the UWFBG hydrophone array, which shows a clear azimuth with much less variation over short periods of time.