Remote Sensing and Noise Processing Methods for Rock Mass Vibration Information in Complex Tunnel Environment
Abstract
During tunnel construction, the surface of surrounding rock is often unstable, making traditional contact sensors (e.g., accelerometers) difficult to install reliably and highly susceptible to detachment or damage. Although scanning laser Doppler vibrometers (SLDVs) offer advantages such as non-contact measurement, high precision, and full-field monitoring, their application for non-contact vibration monitoring in tunnels is subject to numerous constraints. These constraints stem from the specific complexities of tunnel construction environments and the inherent limitations of laser vibrometry technology. To address the challenges of acquiring valid vibration information using laser vibrometers, this chapter presents breakthroughs in both hardware and algorithms. In terms of hardware, an adaptive vibration-damping and noise-reducing carrier platform was developed. This platform mitigates noise interference through physical passive isolation and active control technologies. Furthermore, based on the characteristics of tunnel environments, the scanning laser vibrometry monitoring equipment was optimized for complex tunnel conditions. In terms of algorithms, a hybrid noise reduction method integrating AVMD and K-SVD was proposed. By leveraging complementary enhancement, multi-dimensional analysis, and robustness improvement, this method achieves noise separation and signal enhancement, thereby significantly improving the signal-to-noise ratio (SNR) and accuracy of micro-vibration measurements. The effectiveness of this noise reduction approach was verified through the analysis of simulated signals and measured data, successfully enabling the extraction of valid vibration information from rock masses.
Keywords
Citation Information
@article{chuanyima2026,
title={Remote Sensing and Noise Processing Methods for Rock Mass Vibration Information in Complex Tunnel Environment},
author={Chuanyi Ma and Xuefeng Han and Ning Zhang and Hongliang Liu and Guangyu Yang and Wenfeng Tu and Changyuan Chen and Yuxue Chen},
journal={Scientific Reports},
year={2026},
doi={https://doi.org/10.21203/rs.3.rs-9308153/v1}
}
SinoXiv