Introducing Tree-Based-Regression Models for Prediction of Hard Rock TBM Performance with Consideration of Rock Type

Abstract Prediction of machine performance is a fundamental step for planning, cost estimation/control and selection the type when using tunnel boring (TBM). Penetration rate (PR) utilization (U) are two principal measures TBM evaluating feasibility in given ground condition. However, despite widespread use TBMs established track records, accurate estimation could still be challenge, particularly complex geological conditions. Since different types rocks have varied texture (cementation grain size), respond differently to cutting forces tunnelling, incorporating effects rock prediction models can improve accuracy estimates. The aim this study was develop predicting penetration hard based on field index (FPI), multivariable regression analysis learning algorithm, including classification tree (CART). proposed offer estimated FPIs types, strength, mass properties form graphs (diagrams), which used estimate rate. been developed comprehensive database various offers more estimates by many key parameters available typical geotechnical reports contract documents. also exhibit sensitivity