Deformation prediction of expansive mudstone tunnel based on SSA-LSTM deep learning model

The tunnel in expansive mudstone is easily affected by the deformation of surrounding rocks during construction, posing a threat to the safety of the construction period. To accurately predict the deformation of surrounding rocks, this paper establishes a deep learning model (LSTM) optimized by the Sparrow Search Algorithm(SSA). Taking a certain under-construction tunnel in Sichuan as the engineering background, deformation monitoring points are arranged inside the tunnel to collect monitoring data on surrounding rock deformation. Monitoring data fromtheDK175+197 cross-section is selected and input into the SSA-LSTM model for prediction. The prediction results show that the prediction accuracy of the SSA-LSTM model is better than that of the LSTM model and traditional machine learning models optimized by SSA, with mean square error (MSE), mean absolute percentage error (MAPE), and coefficient of determination (R²) being 0.0118, 2.2757%, and 0.9669 respectively, which are all optimal to meet the requirements for guiding tunnel construction.