Structural stability detection in tunnel pipe jacking based on GWO-PSO

Abstract

In tunnel pipe jacking construction, the complex and variable environment affects the stability and support of the lining structure, which hinders the normal progress of tunnel construction. Therefore, this study proposes a structural stability detection model that integrates the Grey Wolf Optimizer and Particle Swarm Optimization. This model utilizes the preprocessing capabilities of the cubic spline interpolation and wavelet denoising methods, combined with the fast learning ability of the Extreme Learning Machine neural network, to extract structural stability. The results of the study show that the model achieved an Area Under the Curve value of 0.8981, a prediction accuracy of 94%, and stabilizes within 20 iterations. In cumulative data prediction based on the time sequence superposition principle, the maximum offset is 0.055 mm, with the most concentrated absolute error points and a median absolute error of 0.11 μm, outperforming other algorithms. In actual detection, after data processing by this model, the average Signal-to-Noise Ratio increases by 18.57%, the correlation coefficient reaches 0.96, and the standard deviation is 0.019. These results indicate that the model balances computational efficiency and accuracy, demonstrating strong stability and practicality. It improves the efficiency and accuracy of structural stability detection in pipe jacking construction, ensuring smooth construction and safeguarding life and property.