Method for tunnel cross-section deformation monitoring based on distributed fiber optic sensing and neural network

•In this study, a novel method of tunnel cross-section deformation monitoring based on DFOS and a neural network is proposed. The proposed method has the following advantages over the previous methods:•• Based on a neural network algorithm, the automatic search for the mathematical relationship between DFOS strain measurements and the shape curve of the tunnel cross section is realized, which otherwise requires complex mathematical calculation, and thus save considerable computation time.•• When the neural network is trained, the shape curve of the tunnel cross section after deformation can be obtained instantly according to the given DFOS monitoring data.•• As a result, the shape curve of the tunnel cross section is directly mapped from the entire measured strain curve, which prevents the accumulation of measurement error with increasing distance of monitoring.