Research On Double Shielded TBM Performance Prediction Based On Field Measured Data

Compared with the traditional drilling and blasting method,tunnel construction using the TBM method has many advantages such as fast advance rate,low economic cost,safe construction operation,and small impact on the surrounding environment.Priority is given to TBM construction,so this also provides a broad platform for the rapid development of TBM tunnel construction technology.How to improve the TBM performance on the premise of ensuring safe construction is a hot spot in the field of TBM construction.In view of the complexity of the geological conditions and the randomness of the construction process,how to predict the TBM performance scientifically is a difficult problem to be solved in TBM construction.At present,there are still some deficiencies in the research of TBM performance prediction.The parameters selection,prediction methods and applicability of the model need to be further improved.Based on this,this article relies on the National Natural Science Foundation of China's General Project(4192270)and the independent research and development project(2016-ky56(2))of the YREC,taking the Lanzhou water source construction project water delivery tunnel project as the engineering background,and the TBM performance prediction as the research object,making full use of surrounding rock properties,driving parameters,cutter wear and other on-site measured data,combining mathematical statistics with intelligent optimization and other methods to predict the TBM's three performance indicators such as the penetration rate,equipment utilization rate,and advance rate.The research results have important reference significance and engineering guidance significance for argumentation selection of construction method,TBM selection optimization,tunnel construction period prediction and cost budget.The main research contents and conclusions are as follows:(1)Through the single-factor correlation analysis between the TBM penetration rate and various influencing factors,the most important influencing factors of the TBM penetration rate were identified,and five parameters used to predict the TBM penetration rate in this study were determined: Rock uniaxial compressive strength(UCS),rock Brazilian tensile strength(BTS),rock cherchar abrasivity index(CAI),cutter head thrust force(TF),cutter head revolution per minute(RPM).(2)Combining the two methods of partial least squares regression(PLSR)and BP neural network,a prediction model of TBM penetration rate is established,and it is compared with the prediction models established by other methods.The research results show that the prediction model of the TBM penetration rate based on the PLSRBP neural network method avoids the shortcomings of the above two methods alone.The model has the advantages of fast convergence,stable solution and high fitting accuracy.It provides a new method for the prediction of TBM penetration rate.(3)The rock mass rating(RMR),rock cherchar abrasivity index(CAI),and rock hardness(H)are comprehensively considered,and a multivariate nonlinear regression analysis method is used to establish the rock mass correlation utilization rate(Ur).The empirical equation is predicted,and the validity and reliability of the prediction model are verified by other measured data.(4)Based on the uncertainties and risks in the construction of TBM,comprehensively consider the three parameters of rock mass rating(RMR),TBM working condition grade number(TWCR)and risk index(RI),and the prediction model of TBM advance rate was established by BP neural network and the rationality of the model was verified.The research results show that it is unreasonable to predict the advance rate with conventional rock parameters and tunneling parameters,and the prediction results are not accurate enough.The three parameters,RMR,TWCR,and RI,have relatively good correlation with the advance rate.It is feasible to predict the advance rate of TBM.