Wide-frequency-range Vibration Positioning Based on Adaptive TQWT for Long-Distance Asymmetric Interferometer Sensor
Xinxin Hu,Kun Liu,Xibo Jin,Dongqi Zhang,Chaoxiang Chen,Tianhua Xu,Junfeng Jiang,Tiegen Liu
DOI: https://doi.org/10.1016/j.optlaseng.2023.108000
IF: 5.666
2024-01-01
Optics and Lasers in Engineering
Abstract:The asymmetric dual Mach-Zehnder interferometer (ADMZI) based vibration sensor effectively extends the sensing distance, however, the asymmetry seriously deteriorates the positioning accuracy. In this paper, an adaptive tunable Q-factor wavelet transform (TQWT) based long-distance asymmetric sensing system is proposed and experimentally demonstrated for the positioning of wide-frequency-range knocking vibration signals. The TQWT can extract the time-frequency features of the non-stationary signals by performing multi-scale decomposition. Firstly, the positioning method adaptively determines the decomposition levels by analyzing the power spectrum of the vibration signals, which not only effectively suppresses the effect of low-frequency noise, but also accurately extracts the main time-frequency variation characteristics of the vibration signals with various bandwidths. Then, the time delay between the time-frequency characteristics is obtained using a crosscorrelation algorithm, and the vibration position is demodulated. Experimental results show that the method can accurately locate vibration signals with bandwidths of 10-80 kHz at a sensing length of 125 km. For highfrequency, strong vibration signals, the standard deviation of the positioning is around 35.1 m. For lowfrequency, weak signals, the proposed method can still achieve effective positioning compared with the traditional approaches with a standard deviation of approximately 238.1 m. The proposed scheme can significantly improve the applicability of the asymmetric interferometer based vibration sensing system.