基于水化学特征的聚类分析对矿井突水源判别

Prediction of Mine Water Inrush Sources Based on Cluster Analysis of Hydrogeochemical Features

为了能够准确并快速地判别煤矿突水水源位置,基于水化学特征的模糊聚类分析,对突水水源进行综合评判。首先利用突水点的水化学分析指标离子的检测值与以往收集的各个主要含水层的水化学指标离子本底值通过软件绘制出三角形水质图,对比分析可排除最不可能的充水含水层;在此基础上再利用模糊数学聚类分析方法,将剩下的不确定充水含水层建立评判集合进行聚类分析,计算隶属度,通过对计算结果的分析,最终判定突水水源。分析结果表明,采用这2种方法判别突水水源,方法简单易操作,节省时间,可以避免由于水文地质条件的复杂性和模糊聚类分析的模糊性而造成的原因不唯一或结果失真,从而大幅提高判别结果的正确性。

In order to correctly and rapidly predict the location of the mine water inrush sources, a comprehensive judgment was conducted on the mine water inrush sources based on the fuzzy cluster analysis of the hydrogeoehemical features. First, the ion detection value of the hydrogeoehemical analysis indexes of the water inrush location with the previous collected ion background value of the hydrogeochemical indexes of each main water bearing stratum was applied to map the delta water quality drawing and the comparison analysis could remove the most impossible water filled water bearing strata, based on the circumstances, the fuzzy mathematical cluster analysis method was reapplied to the cluster analysis on the judgment assemblage set up with the rest uncertain water filled water bearing strata to calculate the subordinate degree. With the analysis on the calculation results, the inrush water sources could finally predicted. The analysis results showed that those two methods applied to predict the inrush water sources would be easy in operation and saving in time and also could eliminate the not only cause occurred by the complexity of the hydrological conditions and the fuzziness of the fuzzy cluster analysis, which could highly improve the correctness of the prediction results.