Hilfe beim Zugang
Analysis of quartz content in muck based on artificial intelligence algorithms and laser-induced breakdown spectroscopy in TBM tunneling
Abstract Rapid and in-situ sensing of the geological conditions of the rock mass is crucial to the tunnel boring machine (TBM) excavation process. The changes in rock strength and abrasiveness directly affect the working status of the TBM cutters and the tunneling efficiency. The content of quartz,...
Ausführliche Beschreibung
Abstract Rapid and in-situ sensing of the geological conditions of the rock mass is crucial to the tunnel boring machine (TBM) excavation process. The changes in rock strength and abrasiveness directly affect the working status of the TBM cutters and the tunneling efficiency. The content of quartz, as one of the important mineral components of rock, has a direct correlation with the wear state of the cutters. However, the current analysis of rock minerals still lacks a timely and convenient calculation method. In this study, a quartz content prediction method combining laser-induced breakdown spectroscopy (LIBS), rock chemical analysis and artificial intelligence algorithm is proposed. The spectral intensities of the main elements of the rock are obtained by LIBS detection of the muck on the belt machine. Then the quantitative characterization relationship between rock elements and mineral composition is analyzed to establish a transformation matrix. Finally, the mapping relationship between rock major element content and spectral intensity is established based on the random forest algorithm of Bayesian optimization (BO-RF). Subsequently, a prediction model for the mineral content of rocks, including quartz, is established. Relying on the Zhujiang Delta Water Resources Allocation Engineering, the prediction effects of this method and the other two algorithms (BO-Catboost and PLSR) are compared and verified. The results show that the prediction error of quartz content is less than 13% for all three methods, with the smallest error of 10.79% for the proposed method in this study. This also confirm the reliability and engineering practicability of the method. Ausführliche Beschreibung