An Approach to Rapidly Evaluating Rock Mass Quality in Underground Engineering Based on Multi-source Heterogeneous Data

A fast, accurate and easily accessible rock mass quality evaluation method is important for the advancement of underground engineering. In this study, we first analyze the application of mainstream single-factor and multi-factor evaluation methods through extensive tests conducted on the tunnel construction site. Next, by analyzing the evaluation methods for rock mass quality and the selection of evaluation indicators in various standards, we select the most suitable evaluation indicators for the current project from multi-source heterogeneous data based on the principles of scientificity, sensitivity, accessibility, and easy parameterized characterization. After that, we construct a Gaussian process classification model for underground engineering rock mass quality evaluation based on multi-source heterogeneous data. The results show that the accuracy of the tunnel geological prediction method considering single factors is relatively low—only 42.9% for class-V surrounding rocks. The multi-factor evaluation method features simple operation and high accuracy in practical engineering, but its application is limited by lithology. The evaluation method based on multi-source heterogeneous data screening and Gaussian process is free from the influence of geological conditions, and its evaluation results are completely consistent with the actual situation. It can also help provide probability calculation to characterize the uncertainty of underground engineering rock mass quality. The proposed principle for selecting indicators can be used as a reference for the studies in other fields. The proposed evaluation method features high accuracy, simple operation and good applicability.

This study analyzes the applicability of tunnel seismic tomography (TST) to the evaluation of rock mass quality.