中文摘要：

迅速准确地识别矿井充水水源是防治矿井水害事故的重要保证。对未确知聚类方法进行优化,综合考虑水化学指标对水源识别的重要性,选取Na＋＋K＋、Mg2＋、Ca2＋、HCO-3、Cl-、SO2-4、TH等7项水化学指标作为水源识别的判别指标,以新庄孜煤矿22组水样数据作为训练样本,建立了矿井充水水源识别的EW-UCA模型。随后利用该模型对5组测试样本水样进行识别检验并与实际情况作了对比分析。为进一步检验识别模型的可靠度,利用该模型对谢一矿充水水源进行了识别研究。结果表明,EW-UCA水源识别模型具有较高的精确度和较强的适用性,可以在工程实际中推广应用。

英文摘要：

Identifying the water-filled source of mine with reasonable accuracy and with rapidity is the important guarantee of prevention and control of mine water disasters. In the paper,uncertainty measurement theory is optimized. Seven hydrochemistry indexes including Na＋＋ K＋,Mg2 ＋,Ca2 ＋,HCO-3,Cl-,SO2-4and TH are regarded as the discriminant index for identifying water source,combined with the importance of hydrochemistry index to water source＇s identification. The EW-UCA model for identifying mine＇s waterfilled source is established,based on 22 groups of water samples as the training sample of Xinzhuangzi Coal Mine. Followed by 5 groups of water samples,the testing samples are identified and compared to the actual situation by the model. In order to further test the model＇s reliability,the model is used to identify the waterfilling source of the Xieyi Coal Mine. The results show that the EW-UCA model for identifying water-filling source is reliable,and practical and can be applied to the actual engineering.