综合灰色系统理论与传统的边坡岩体质量分级方法(SMR法)，提出改进SMR法。传统的岩体质量分级方法中定量指标取值离散性很大，造成质量分级结果阶梯变化。灰色系统理论的灰度特征对解决这类小样本、离散性的问题有很好的适用性。首先对传统质量分级方法的评价指标进行灰类划分，确定各指标所占权重，再构建评价指标的三角白化权函数，并基于最大隶属度准则对边坡岩体进行质量分级。最后结合工程边坡实例，与一般工程RMR(岩体质量分级)与SMR法比较，改进SMR法的评价结果更加吻合工程现状，且质量分级稳定性高，表明其应用于边坡岩体质量分级是科学和准确的。

The improved slope mass rating method(SMR) is proposed, through integrating the gray system theory with traditional SMR method. The quantitative indexes value of traditional rock mass classification methods was very discrete, resulting in the step change in classification. The problems of small samples and discrete issues are solved by using features of gray system theory. Firstly, the evaluation indexes of traditional rock mass quality classification are classified. Also, the weights of indexes are determined; and the triangle whiten function for evaluation indexes is built. Furthermore, the rock mass quality classification of slopes is valued based on the maximum degree. Finally, combining with a slope case study and comparing general engineering rock mass classification methods(RMR and SMR), the result judged by improved SMR method can agree well with engineering facts; and it expresses a high and strong quality grade stability. For rock mass quality classification in slope engineering, this study shows that this method is scientific and accurate.