Research on Tunnel Surrounding Rock Displacement Prediction Based on Deep Learning
The prediction and forecast of tunnel surrounding rock deformation has important theoretical and practical significance for guiding tunnel construction,dynamic protection,and ensuring long-term safe operation of tunnels.To more effectively predict the deformation of tunnel surrounding rock,taking the measured arch displacement monitoring data of a tunnel in highland as an example,the variational mode decomposition(VMD)method was introduced for monitoring time series preprocessing,extracting and decomposing key deformation signals of surrounding rock.A deep learning prediction of decomposition displacement using a gated recurrent unit(GRU)model and a long short-term memory(LSTM)model optimized by particle swarm optimization(PSO)algorithm was proposed.Furthermore,the prediction results of all sub models were merged to reconstruct the cumulative displacement prediction results of arch crown settlement.The example verification results show that the VMD-PSO-GRU model proposed in this article can improve the prediction accuracy of tunnel surrounding rock deformation,and serve as a new method for predicting tunnel surrounding rock deformation.
NETL
NSTL
万方数据
