Research On Classification Of Rockburst Intensity And Criterion Of Stress Intensity For Lasa To Linzhi Railway Tunnel

In tunnel engineering with high ground stress and hard rock surrounding rock,rockburst is a common and huge hazard.Since the geostress field and surrounding rock conditions across the entire tunnel are complex,the prediction methods proposed by the predecessors based on the rockburst of a certain tunnel or the rockburst record of a certain area of the tunnel are often not applicable in a new tunnel with a different situation.It can be concluded that there is no one-size-fits-all method in the study of rockburst prediction.The Sichuan-Tibet Railway is a main railway connecting the Sichuan-Tibet area.Many of the tunnels are in high ground stress and hard rock surrounding,and have a high risk of rockburst,therefore rockburst prediction is particularly important.However,in some completed tunnel projects,the accuracy rate of rockburst prediction is not high.In view of this,this article relies on the Sangzhuling Tunnel Project of the Lalin Section of the Sichuan-Tibet Railway and proposes a rockburst intensity classification standard and a rockburst criterion,with a view to providing a reference method for rockburst identification and prediction for the Lalin Section and even the Sichuan-Tibet Railway.The main research work and research results are as follows:(1)Through systematic data investigation,based on the Sangzhuling tunnel from Lasa to Linzhi,the method and principles for formulating the rockburst intensity plan were established,the rockburst field measurement in the research section of the tunnel was conducted,the rockburst intensity classification plan based on the statistics of the measured results was proposed,and the classification of the intensity levels of rockburst cases that have occurred in the study section of the tunnel was completed.(2)Based on the research of existing rockburst prediction methods,the criterion form of stress-intensity ratioσ_θ/σ_c suitable for Sangzhuling Tunnel was selected,the value ofσ_cwas determined by field point load test,and the field stress relief method and a numerical model is established to determine the value ofσ_θ.Then the k NN machine learning algorithm is used to obtain the threshold value ofσ_θ/σ_c corresponding to the rockburst intensity level,and a rockburst criterion in the form of stress intensity criterion prediction is proposed.(3)Based on the on-site rockburst measurement records of the Dagala Tunnel,Zhulagang Tunnel,and Gangmula Tunnel from Lasa to Linzhi,a numerical calculation model was established,the value ofσ_θ/σ_c of the target section was calculated,and the rockburst criterion proposed was used for rockburst.For rockburst prediction results,the similarities and differences between the proposed rockburst criteria and the existing rockburst criteria of the same type are analyzed and evaluated.(4)Based on the basic idea of the machine learning method,the improved form of the rockburst criterion was discussed.On one hand,a method for improving the threshold value of the rockburst criterion based on updated training samples was established,and the k NN machine learning algorithm was used to propose an improved threshold of the rockburst criterion;on the other hand,a comprehensive prediction method based on considering multiple factors is proposed,the applicability of the XGBoost algorithm in the field of rockburst prediction under the small sample case is studied,and a rockburst prediction model based on the XGBoost algorithm is established.