基于组合赋权和未确知测度的深埋隧道岩爆危险性评价——以川藏交通廊道桑珠岭隧道为例

针对复杂山区深埋隧道岩爆危险性评价中的诸多不确定性因素问题，通过归纳分析典型高地应力条件下深埋隧道岩爆破坏特征及关键影响因子，从客观反映高地应力环境、岩石力学性能和围岩性质3个层面确定5项岩爆评价指标，利用未确知测度理论建立隧道岩爆危险性评价模型.为了充分考虑岩爆危险性评价的主观因素和客观因素，通过引入距离函数，采用熵权法和层次分析法相结合构建组合赋权法，综合确定各指标的权重系数.基于未确知测度理论及计算规则，结合岩爆危险性分级标准，构建直线型单指标测度函数，计算单指标测度评价矩阵和多指标测度向量，依照置信度准则进行岩爆危险性评价.将构建的岩爆危险性评价未确知测度模型应用于川藏交通廊道桑珠岭隧道，并与强度应力比法、Russenes判据、岩石脆性系数、岩体完整性系数、岩石弹性能指数等单指标判据评价结果及实际岩爆结果进行对比.研究结果表明:该模型评价结果的准确率达到94.4%，比单指标岩爆判据的准确率高16.7%~66.7%. 

Rockburst Risk Assessment of Deep Lying Tunnels Based on Combination Weight and Unascertained Measure Theory: A Case Study of Sangzhuling Tunnel on Sichuan-Tibet Traffic Corridor

Aming at the uncertain factors in the rockburst risk assessment of deep lying tunnels in complex mountainous areas, the unascertained measurement theory was used to establish a tunnel rockburst risk evaluation model. By summarizing and analyzing the rockburst failure characteristics and key influencing factors of deep lying tunnels under typical high geostress conditions, 5 evaluation indexes were determined from three levels that objectively reflect the high geostress environment, rock mechanical properties and surrounding rock properties. In order to fully consider the subjective and objective factors of rockburst risk assessment, by introducing a distance function, using the combination of entropy weight method and analytic hierarchy process to construct a combination weighting method, and comprehensively determined the weight coefficient of each index. Based on the unascertained measurement theory and calculation rules, it constructed a linear single-index measurement function according to the rockburst risk classification standard. By calculating the single-index measurement evaluation matrix and the multi-index measurement vector, the rockburst risk assessment was carried out according to the confidence criterion. The unascertained measurement model of rockburst risk assessment was applied to the Sangzhuling tunnel of the Sichuan-Tibet traffic corridor. The evaluation accuracy was compared with that obtained from single index criteria such as strength-stress ratio method, Russenes criterion, rock brittleness coefficient, rock mass integrity coefficient, rock elastic energy index and the actual rockburst results. The research results show that the accuracy rate of the evaluation result of the model is 94.4%, which is 16.7%-66.7% higher than that of the single-index rockburst criterion.