论文部分内容阅读
采用熵权法和云模型判定岩爆等级。选用岩石的单轴抗压强度σ_c、单轴抗拉强度σ_t、切向应力σ_θ、岩石的压拉比σ_c/σ_t、岩石的应力系数σ_θ/σ_c和岩石的弹性变形指数W_(et)作为岩爆等级判定的因素建立岩爆评价指标体系。以收集到209组工程中的实际岩爆情况及数据作为样本进行分析计算,建立岩爆等级判定的熵权-云模型。运用该分析模型分析岩爆评价指标体系中评价指标的敏感性,并对收集到的工程实例岩爆情况进行判定,将结果与Bayes、KNN和随机森林方法的判定结果进行比较。研究表明:评价指标体系中指标敏感性由大到小的顺序为:σ_θ /σ_c, σ_θ, W_(ct), σ_c/σ_t, σ_t, σ_c,;熵权-云模型的判别准确率比Bayes、K最邻近结点算法(KNN)和随机森林(RF)方法高。
Entropy Method and Cloud Model to Determine Rock Burst Level. The rock uniaxial compressive strength σ_c, uniaxial tensile strength σ_t, tangential stress σ_θ, rock tensile ratio σ_c / σ_t, rock stress coefficient σ_θ / σ_c and rock elastic deformation index W_ (et) Blast level to determine the factors to establish rockburst evaluation index system. Taking the actual rockburst situation and data collected from 209 groups of projects as samples, the entropy-cloud model of rockburst grade determination is established. The analysis model was used to analyze the sensitivity of the evaluation indexes in the rockburst evaluation index system. The rockburst situation of the collected engineering examples was judged, and the results were compared with those of the Bayes, KNN and random forest methods. The results show that the order of the sensitivity of the indicators in the evaluation index system is: σ_θ / σ_c, σ_θ, W_ct, σ_c / σ_t, σ_t, σ_c, and the discriminant accuracy of the entropy-cloud model is better than Bayes, K Nearest Neighbor Algorithm (KNN) and Random Forest (RF) methods are high.