Displacement prediction method of underground excavation based on self-memorization model of dynamic system

The self-memorization principle of dynamic system was introduced into the deformation prediction research of surrounding rock. By treating the monitored displacement time series data as the particular solution of the nonlinear dynamic evolvement model of surrounding rock in underground excavation, the nonlinear ordinary differential equation of surrounding rock deformation was deduced based on bilateral difference principle. The self-memorization model for displacement prediction of surrounding rock in underground excavation was established on the basis of the deduced nonlinear differential equation. The model was applied to model and predict the deformation time series data monitored at 505 m haulage drift in Tongkeng Stannum Mine and the outlet tunnel in Shuibuya Hydropower Station. The engineering case study shows that the self-memorization model has good modeling and prediction accuracy, has good applicability for different displacement-time curves of underground excavation.