Real-time prediction of mechanical behaviors of underwater shield tunnel structure using machine learning method based on structural health monitoring data

Predicting the mechanical behaviors of structure and perceiving the anomalies in advance are essential to ensuring the safe operation of infrastructures in the long run.In addition to the incomplete consideration of influencing factors,the prediction time scale of existing studies is rough.Therefore,this study focuses on the development of a real-time prediction model by coupling the spatio-temporal correlation with external load through autoencoder network(ATENet)based on structural health monitoring(SHM)data.An autoencoder mechanism is performed to acquire the high-level representation of raw monitoring data at different spatial positions,and the recurrent neural network is applied to understanding the temporal correlation from the time series.Then,the obtained temporal-spatial information is coupled with dynamic loads through a fully connected layer to predict structural performance in next 12 h.As a case study,the proposed model is formulated on the SHM data collected from a representative underwater shield tunnel.The robustness study is carried out to verify the reliability and the prediction capability of the proposed model.Finally,the ATENet model is compared with some typical models,and the results indicate that it has the best performance.ATENet model is of great value to predict the realtime evolution trend of tunnel structure.

Assessing site investigation program for design of shield tunnels

In the design of shield tunnels,it is expected that a design can be more economical as more site investigation data are available.Nev-ertheless,the cost of site investigation will also increase as more site investigation data are required.It is thus important to assess the potential benefit of a site investigation program before it is conducted considering the uncertainty in the soil properties and the site inves-tigation outcomes.In this paper,a probabilistic framework is suggested to assess the effectiveness of a site investigation program for the design of shield tunnels through the random field theory.An efficient method based on the generalized extreme value distribution is used to calculate the failure probability of the tunnel,through which the expected benefit from a site investigation program can be estimated conveniently.The result shows that a site investigation program is more valuable when a greater target design reliability index is needed.The expected benefit from a site investigation program increases as the borehole intensity increases,and it also increases as the scale of fluctuation of the soil properties increases.The method suggested in this paper provides a useful tool for planning borehole layout for the design of shield tunnels,which is promising for the optimization of site investigation.