Method for shield tunnel cross-section curve reconstruction based on FBG sensors and neural network

Abstract

To improve the accuracy of reconstructing morphology in shield tunnel monitoring by fiber-optic Bragg grating (FBG) sensors, a Transformer network-based optimized tunnel cross-sectional curve reconstruction method is proposed to solve the problem of low reconstruction accuracy due to the cumulative error of discrete curvature points. In this research, the optimization method uses all interpolation points as initial points to minimize the cumulative error in fitting the curve with discrete curvature points. Simulation model experiments of the shield tunnel were carried out to verify the accuracy of the optimization method in reconstructing the tunnel section curve. The results show that the shield tunnel simulation model experiments, more accurate tunnel section reconstruction curves were obtained using this optimization method, increasing reconstruction accuracy by 27.3424%.