A Case Study of Visualization Prediction of Deformation of a Typical Rock Tunnel Using Variable Modal Decomposition Technique, Memory Networks, and BIM Technique

Abstract

A visual deformation prediction method was proposed to improve the accuracy and visualization of the surrounding rock deformation prediction in tunnel construction, combining the Variational Modal Decomposition (VMD) and Bidirectional Long- and Short-Term Memory (BiLSTM) network. Based on the VMD method to decompose the measured data of tunnel surrounding rock deformation, the BiLSTM model was used to predict the final deformation value. The prediction results were also embedded into the tunnel’s Building Information Modeling (BIM) as plug-ins, and the data were visualized through graphs and color warnings. Taking the measured data of the arch settlement of the Loushan tunnel as an example, the results showed that the prediction results were more consistent with the measured situation, and the visualization expression could effectively warn of the risk of vault settlement in the construction stage. This study realized the combined use of surrounding rock deformation prediction and BIM technology, which could be used as a reference for similar projects.