To address the problem of early warning of pipeline leakage, this letter proposes a pipeline leakage early warning method based on DAS. Through DAS monitoring the vibration signals induced by different degree of pipeline corrosion and blockage, using pattern recognition techniques, distributed early warning and location of pipeline leakage are realized. By conducting simulations and experimental analyses, the differences in vibration characteristics caused by four types of pipeline anomalies were validated. Combining wavelet packet energy ratios and the BP neural network, the average recognition rate for the five event classes (normal pipeline, 40% corrosion, 80% corrosion, 40% blockage, 70% blockage) was 95.39% based on 3910 samples. The results demonstrate that the proposed method can effectively achieve the early warning of pipeline leakage.
4 August 2023 in Measurement Science and Technology
The integration of artificial intelligence and distributed fiber optic acoustic sensing technology (DAS) has yielded remarkable results in recent years; however, some application scenarios face the challenge of acquiring an adequate amount of data for higher network accuracy. To address this issue, we propose a decoupling parallel convolutional neural network (DPCNN) that relies on multiple feature inputs to achieve higher accuracy while using smaller databases. Our model offers excellent recognition of five events, including background noise, footstep, digging, car passing, and climbing fence, with an accuracy rate of up to 94.9%. The DPCNN is a parallel and lightweight convolutional neural network (CNN) that boasts a short training time of only 3.76 s per epoch and a test time of 0.1175 s, with superior network convergence. In comparison to a mature single-branch CNN based on mixed images of time-frequency and time-space, the DPCNN accuracy is 6.4% higher. Our model demonstrates excellent performance across various databases and can achieve recognition accuracy of up to 98.7% with larger databases. Finally, we show the broad range of applications available for DPCNN based on multiple feature inputs when using a mature single-branch replacement in each branch of a two-branch network.
We demonstrate a fiber Bragg grating (FBG) array based wavelength calibration scheme for Fourier domain mode-locked (FDML) laser. The wavelength interval and the temperature feedback module of the FBG array are designed to ensure the reference stability of the wavelength calibration scheme. Combined with the calibration scheme, the FDML laser with a tunable wavelength range of ∼60 nm, a center wavelength of 1300 nm and a sweep frequency of 39.63 kHz is built up to demonstrate its feasibility. The FBG wavelength demodulation based on the calibrated FDML laser system shows a wavelength resolution of 2.76 pm and hourly stability of 10.22 pm.