Intelligent fine-grained classification of tunnel surrounding rock mass by integrating tunnel face point clouds and measurement-while-drilling data

Fine-grained rock mass classification is essential for support design and safety control in drill-and-blast tunnels. To address the limitations of conventional manual geological logging, including subjectivity, delayed feedback, and difficulty in continuous quantitative assessment, this study proposes an intelligent fine-grained rock mass classification method that integrates tunnel face point clouds and measurement-while-drilling (MWD) data from drilling jumbos. Tunnel face point clouds are used to identify the development characteristics of discontinuities and quantify rock mass integrity, while MWD parameters are employed to characterize the relationship between drilling responses and rock mass conditions. The structural geometric information and drilling-induced mechanical responses are further mapped into a unified probabilistic space for fusion-based classification, thereby improving the accuracy and reliability of rock mass classification. Engineering validation was conducted using typical construction cycles of a drill-and-blast tunnel. The results show that the proposed method can rapidly distinguish between Grade IV and Grade V rock masses, with the fusion-based classification results consistent with manual judgments. The complete workflow for a single construction cycle can be completed in approximately 12–19 min, meeting the requirements for rock mass classification and support decision-making during construction intervals. The findings provide an interpretable and deployable technical approach for intelligent construction and dynamic risk control in drill-and-blast tunnels.