Due to the complexity,suddenness,and diversity of forms of rockburst disasters,issues like accurate perception and warning of disaster sources and precise control of stress fields urgently need to be addressed.Digital drilling is considered a new in-situ testing method for quantitatively measuring rock mechanics information.An experimental system with triaxial high-stress loading for drilling simulation was innovatively developed.With the aid of this experimental system,drilling sensitivity experiments were conducted under different stress and drilling conditions to investigate the influences of different con-ditions on various drilling information,including torque,thrust,and drill cuttings.Additionally,a me-chanical model for identifying coal seam stress while drilling was established,and the mechanism of cor-relation between coal seam stress and drilling response information was revealed.On this basis,a multi-parameter-fused intelligent identification method for coal seam stress based on back propagation neural network(BPNN)was proposed and applied to engineering practice.The following findings were ob-tained.Drilling information(torque,thrust,et al)is positively correlated with loading stress,and its growth trend is identical to the growth trend of drill cuttings with stress.Drilling speed and rotational speed are important factors affecting drilling response information.The BPNN-based identification model for coal seam stress while drilling achieves an excellent fitting effect,its prediction accuracy being as high as 95%.According to the application to underground engineering practice,the stress-anomalous area obtained through active computed tomography(CT)is highly consistent with the stress obtained by inversion while drilling,which verifies the accuracy of identifying coal seam stress while drilling.This study,which provides a scientific basis for the implementation of in-situ digital drilling technology for coal seam stress,is beneficial for achieving intelligent perception and early warning of rockburst hazards and promoting the development of efficient and intelligent rockburst prevention technology.
关键词
深部开采/冲击地压/应力随钻监测/三轴试验/神经网络
Key words
deep mining/rockburst/stress monitoring while drilling/triaxial test/neural network
维普
万方数据