Automated Monitoring of Structural Deformation Signals in Overlying Subway Tunnels Induced by Large-Section Underground Excavation

Abstract Large-section underground excavation is one of the most common operations in subway tunnel construction, which is highly prone to causing tunnel deformation. To address this issue, an automated monitoring method for structural deformation signals in overlying subway tunnels induced by large-section underground excavation is proposed. The method employs wavelet transform to decompose subway tunnel signals into three layers, reconstructs the signal structure through threshold quantization, and achieves denoising of subway tunnel structural signals. By utilizing wavelet functions and scale factors, the interval of structural deformation signals is determined. The presence of deformation signals within the monitoring area is identified based on the maximum likelihood estimation of adjacent signal parameters. The disturbance coefficient is introduced to narrow the range of abnormal signals, enabling automated monitoring of subway tunnel structural deformation signals. Experimental results demonstrate that the proposed method, after incorporating the disturbance coefficient, accurately identifies deformation signals, reducing the range of abnormal signals by 10%. The method exhibits a high signal-to-noise ratio, with the RMSE value stabilizing around the optimal value of 0.002. The monitoring process is stable, with a consistent ratio between time and monitoring points across each time period. It has been proven that the author's method is highly efficient, accurate, and stable in the automated monitoring of structural deformation signals in subway tunnels.