矿井突水水源判别的多组逐步Bayes判别方法研究

Studies of water source determination method of mine water inrush based on Bayes' multi-group stepwise discriminant analysis theory

矿井突水是采矿过程中最具威胁的自然灾害之一。能否准确快速地判别矿井突水水源,不仅是矿井水文地质工作的主要内容,而且是煤矿防治水工作的重要基础。为了有效判别矿井突水水源,综合考虑水化学指标对水源判别的重要性,基于多组逐步Bayes判别分析理论,选取Na++K+、Ca2+、Mg2+、Cl-、SO42-和HCO3-共6项指标作为判别因子,利用国内某矿区各主要含水层的35个水样的水质资料作为训练样本,建立了矿井突水水源预测的多组逐步Bayes判别分析模型。实例分析表明,该模型结果与实际情况相符合,说明该模型在矿井突水水源判别中具有良好的实用性和有效性,为判别矿区新的突水水源提供了一种新思路。

Mine water inrush is one of the most threatening disasters in mining engineering, especially in underground coal mines. How to distinguish the source of mine water inrush exactly and quickly is not only the major content of mine hydrogeology works, but also the important foundation of mine water prevention. In the source determination of mine water inrush, it is essential to consider the importance of the indexes of chemical elements of water. Based on the Bayes＇ multi-group stepwise discriminant analysis theory, six indexes of chemical elements, such as Na^＋ ＋ K^＋, Ca^2＋, Mg^2＋, Cl^-, SO4^2＋ and HCO3^-, are taken as the key factors. The total of 35 different water sources examples from a real mining area are chosen as the training and testing samples, a forecast model for distinguishing the source of mine water inrush is then established. The results obtained by the forecast model in case studies are well consistent with the practical situations, so as to prove the utility and efficient of the method. It may provide a new method for the water source determination of mine water inrush.