LAUSR

Rock mass grading is a basic problem in the construction industry and underground engineering research. Because the index parameters that affect the rock mass quality are ambiguous and random, rock mass quality classification is often uncertain. Based on this issue, this paper selects the rock quality index RQD, rock uniaxial saturated compressive strength Rw, rock mass integrity coefficient Kv, structural surface strength coefficient $K_{f}$ and groundwater seepage quantity $\omega $ as quantitative evaluation indicators to construct an evaluation system. Thirty sets of data collected in China are selected as learning samples. Through the related concepts and finite interval cloud model, the characteristic parameters of the measured data are obtained, and a cloud model is generated with a forward cloud generator to achieve the transformation between qualitative and quantitative concepts. Combined with the basic knowledge of rough set theory, the weight determination problem is transformed into an attribute importance problem. To avoid zero weights in the traditional rough set approach, this paper introduces a calculation method based on the conditional information entropy, and the weight calculation method is modified to obtain the comprehensive weights. According to the principle of the maximum membership degree, the classification of rock mass quality is performed, and the rock mass quality data are determined to have different levels of comprehensive membership. A rock mass quality evaluation method based on the coupled improved rough set–cloud model is established and successfully applied for a rock mass quality evaluation of the 0+000~0+560 test section of the second stage of underground engineering at the Guangdong Pump Storage Power Station. The results show that the model is reliable and practical and provides a new approach for uncertainty analysis and evaluation in rock engineering practice.