Eliminating the Fading Noise in Distributed Acoustic Sensing Data

IEEE Transactions on Geoscience and Remote Sensing, Год журнала: 2023, Номер 61, С. 1 - 10

Опубликована: Янв. 1, 2023

Fading noise is a crucial problem in distributed acoustic sensor (DAS) system which can severely degrade the system's ability on detection. In this paper, we systematically studied mechanism and characteristics of fading dual-pulse heterodyne demodulated DAS system. Results show that not only causes big errors signal, but also leads to amplitude distortion seismic wave. We propose sort average over trace (SAOT) algorithm eliminate data. After implementation simulated data, error be eliminated 100%, residual standard deviation reduced by 32 dB. Meanwhile, correlation between processed data noise-free improved 29 Then performance further verified laboratory experiment, achieving reduction 33 Finally, perfectly vertical profile (VSP) surface obtained at oilfield site.

Язык: Английский

---

Fading Noise Suppression Method of Φ-OTDR System Based on GA-VMD Algorithm

IEEE Sensors Journal, Год журнала: 2023, Номер 23(19), С. 22608 - 22619

Опубликована: Авг. 22, 2023

A distributed optical fiber acoustic sensing system (DAS) based on phase-sensitive time domain reflectometry ( $\Phi $ -OTDR) can detect the vibration information along an and has been widely used for various engineering detections; however, high-coherence detection light source in a DAS will cause coherent fading fiber, leading to noise affecting monitoring of target events. In order solve problem DAS, this study considered method suppressing genetic algorithm (GA) variational mode decomposition (VMD). Furthermore, parameter selection VMD is improved by introducing GA. By sample entropy as fitness function combining it with GA, adaptively searches optimal parameters iterating minimum value according type input signal. Then, decomposes noisy data into multiple intrinsic functions (IMFs), permutation time-frequency each component were extract effective IMF components signal reconstruction obtain denoised Experimental results show that compared empirical (EMD), ensemble EMD (EEMD), complete EEMD (CEEMD) methods, proposed eliminate aliasing problem, effectively suppress improve signal-to-noise ratio, reduce root-mean-square error. It is, thus, possible more accurately monitor events, improving practical application DAS.

Язык: Английский

---

High Precision Raman Distributed Fiber Sensing Using Residual Composite Dual-Convolutional Neural Network

Journal of Lightwave Technology, Год журнала: 2024, Номер 42(10), С. 3918 - 3928

Опубликована: Фев. 15, 2024

Raman distributed optical fiber sensing has the unique ability to measure spatially profile of temperature that are great interest numerous field applications. However, performance is severely limited by signal-to-noise ratio (SNR). The existing SNR enhancement schemes have drawbacks such as increased system complexity, degradation sensor metrics spatial resolution, poor denoising performance, etc. Here, we report residual composite dual-convolutional neural network (RRCDNet), a novel convolutional network-based model for one-dimensional signals specifically tailored sensing. RRCDNet-enhanced dramatically improves precision more than factor 100, from 7.57°C 0.06°C, without hardware modification or other metrics. At same time, RRCDNet can also enhance systems with signals, Rayleigh and Brillouin systems.

Язык: Английский

---

Three-dimensional multichannel waveguide grating filters

Opto-Electronic Science, Год журнала: 2024, Номер 3(12), С. 240003 - 240003

Опубликована: Янв. 1, 2024

Язык: Английский

---

Measuring early-age shrinkage of concrete incorporating shrinkage reducing admixtures using distributed fiber optic sensors

Measurement, Год журнала: 2025, Номер unknown, С. 116763 - 116763

Опубликована: Янв. 1, 2025

Язык: Английский

---