矿井突水水源Logistic识别及混合模型

Logistic identify and mixing model of mine water inrush sources

针对矿井地下水混合度较高的突水水源识别问题,运用多元统计分析原理和混合计算原理,建立焦作矿突水水源识别模型和混合模型.以实际数据作为训练样本,分别对它们进行分析与检验.研究结果表明:Logistic分析能有效建立混合度较低的突水水源识别模型,回估误判率较低;混合模型利用主成分分析分析结果建立四面体,损失较少的信息,可有效地确定地下水的混合比例,且利用示踪元素得到的预测值与实测值的总体误差相对较低.

Aiming at the recognition problem of mine groundwater mixed with high degree of water inrush source, using the multivariate statistical analysis principle and the mixed calculation principle comprehensively, this paper established identification model and mixed model of the Jiaozuo mine water inrush, and analyzed and tested on them respectively based on actual data as training samples. The results show that Logistic analysis can effectively establish a low degree of mixed water inrush identification model with lower error rate. The mixed model uses the principal component analysis, and establishes the tetrahedron which has less information loss and can effectively determine the mixing ratio of groundwater, and use the tracer element to get the predicted value, and the overall error between the predicted value and the measured value is relatively low.