Research On Parameter Identification Of TBM Rock Chips Based On Machine Vision

Tunnel boring machine is now widely used in underground engineering,as a direct product of the excavation process,the shape,particle size distribution,volume and other parameter characteristics of rock chips have important observational significance.At present,manual observation and slag field measurement methods are mostly used for the statistics of their characteristics.In order to solve the hysteresis of rock chips parameter identification and improve the accuracy,a machine visionbased rock ballast parameter identification and measurement system was proposed.Work on image segmentation,shape and size parameter identification and volume measurement of rock chips.The specific research contents are as follows:(1)Based on the analysis of the shape,size,volume parameters of the rock chips,a set of rock chips identification and measurement system is designed,and the overall design and function of the system are introduced.The selection of key components of the data acquisition subsystem is completed,and the camera calibration is carried out.(2)Researched the image segmentation algorithm of rock chips based on deep learning,obtained the shape parameters of rock ballast sheet,and evaluated by the accuracy rate,F1 Score,and IOU index.Based on the point cloud data,two methods for calculating the volume flow of rock ballast pieces based on the TIN model and surface element integration are proposed respectively.(3)A TBM rock chips parameter identification test bench was built to verify the effectiveness of the deep learning model for identifying the shape and size parameters of rock ballast sheets.The results show that the identification results can replace the actual distribution,and the fitting degree of the distribution law can reach 92%;The volume flow measurement experiment is carried out,and the results show that the surface element integration method is more suitable for the volume measurement of rock ballast pieces,the error is within 8%,and the repeatability accuracy is above 95%.(4)The identification method of the characteristic parameters of the rock chips is applied in the TBM project.Combined with the engineering data,the relationship between the rock mass integrity of the face and the rock ballast sheet is analyzed,the index of volume joints is judged by indicators such as inhomogeneous coefficient,curvature coefficient,and block proportion.