为降低采空区稳定性评价结论的不确定性,构建了基于D-S证据理论多源信息融合的采空区稳定性评判模型。利用现场工程地质调查、室内岩石力学实验、数值模拟试验、采空区围岩体声发射监测等方法获取的共20种定量和定性的物理量信息,组成采空区稳定性评价多源证据指标体系。运用欧氏距离法确定各证据指标所属评价等级的质量分配函数,采用多比例两两对比法与熵值法确定各证据指标的权重,通过D-S理论对多源证据指标进行信息融合。根据融合结果计算出信任度函数并以此为依据确定采空区稳定性等级。以盘龙铅锌矿-70m中段5个采空区的稳定性分析为例进行了稳定性评价,评价结论与矿山工程实际相符,验证了多源证据指标体系评价方法的可行性和有效性,为采空区稳定性评价提供了一种新途径。 In order to reduce the uncertainty of the conclusion of the stability evaluation of goaf, a mined-out area stability evaluation model based on multi-source information fusion of D-S evidence theory was constructed. A total of 20 kinds of quantitative and qualitative physical quantities obtained by on-site engineering geological survey, indoor rock mechanics experiment, numerical simulation test and acoustic emission monitoring of surrounding rock mass of goaf are adopted to form the multi-source evidence index system of stability evaluation of goaf . The Euclidean distance method is used to determine the mass distribution function of each evaluation grade of each evidence index. The weight of each evidence index is determined by multiple ratio comparison method and entropy method, and the information fusion of multi-source evidence index is carried out by D-S theory. According to the fusion result, the trust function is calculated and used to determine the stability level of the goaf. Taking the stability analysis of 5 goafs in the middle section of Panlong Lead-Zinc Mine-70m as an example, the stability evaluation is carried out. The evaluation conclusion is in line with the mine engineering practice, which verifies the feasibility and effectiveness of the multi-source evidence index system evaluation method. It provides a new way to evaluate the stability of goaf.