Method And Application Of Geological Research And Parameter Optimization For Large-diameter Shield Tunneling In Soil-rock Composite Strata

With the advantages of high tunneling efficiency and strong safety,shield tunneling machine has become the main mechanical equipment in tunnel and underground engineering construction.However,in tunneling,the control parameters of shield tunneling often do not match and adapt to the formation conditions,which leads to slow advancing speed,shield tunneling stuck machine,mud cake,cutter head abnormal wear,and other engineering problems.There are two main reasons for this.One is that the long-distance blind area will appear in the drilling exploration before the tunnel’s construction,especially the undersea tunnel,due to the large interval between the drilling holes.The formation information between the holes can only be inferred,and the survey report results are quite different from the actual formation conditions,thus misleading the judgment of the construction personnel.Second,it is difficult to adjust the operating parameters timely and reasonably according to the change of stratum conditions because the operating parameters of the shield in the past mainly rely on experience selection.The above engineering problems are more common when shield tunneling through soil and rock composite strata.Therefore,it is of great significance to carry out research on synchronous geological research and parameter optimization methods for large-diameter shield tunneling in soil and rock composite strata to ensure safe and efficient tunneling of shield tunneling.Firstly,for the identification of soil and rock composite strata,this paper constructs a sample data set corresponding to the stratum category and excavation parameters through field tests,adopts Support Vector Machine(SVM)algorithm,improves the algorithm based on Grid Search(GS)and cross-validation,and proposes a GS-SVM based large diameter shield The algorithm is based on Grid Search,GS and cross-validation,and proposes a GS-SVM-based method for classifying strata in excavation,which enables accurate identification of composite strata based on key excavation parameters such as propulsion speed and cutter torque;secondly,it enables identification of composite height ratio of strata based on the difference in rotational speed of disc-type hob in soft and hard strata.The excavation parameters are divided into active and passive control parameters,and a support vector regression(SVR)prediction model is constructed with active control parameters and stratigraphic parameters as input and passive control parameters as output,and a heuristic algorithm,the Grey Wolf Optimizer,is introduced.The heuristic algorithm Grey Wolf Optimizer(GWO)was introduced to optimise the regression prediction model,achieving accurate prediction of passive control parameters such as squeezing pressure and cutter torque.The shield advancement speed is obtained by double constraint calculation.The decision optimisation method was successfully applied to the Shantou Bay Cross-Harbour Tunnel(shield composite stratigraphic section),and the effectiveness and feasibility of the objective optimisation decision method established in this paper was verified by means of data extrapolation and field tunnelling tests.