基于聚类分析的煤矿瓦斯涌出量预测方法研究  

Study on Prediction Method of Coal Mine Gas Emission based on Clustering Analysis

作　　者：姜小龙[1] 娄建成[1] 李福玲[1] 

机构地区：[1]内蒙古科技大学矿业工程学院,内蒙古包头014010

出　　处：《煤》2013年第4期1-2,14,共3页Coal

摘　　要：将影响煤矿瓦斯涌出量因素作为数据对象属性,运用聚类分析对瓦斯涌出量进行聚类,得到各个样本的类别特征及各簇的平均值和标准偏差。并选取其中3个主要影响因素生成三维散点图,分析聚类效果。通过待测样本的类别划分,证实了该方法对煤矿瓦斯涌出量预测的有效性和可行性。Use the impact of coal gas emission factors as data object attribute, and cluster gas flow - volume using clustering analysis, so get all the sample category features, and each cluster of the mean and standard deviation. And they selected three main influencing fac- tors to generate three - dimensional scatter and analyzed clustering effect. Through the data for classification of the sample, demonstrates the validity and feasibility of this method which can predict the volume of mine gas emission.

关 键 词：聚类分析 K—means 瓦斯涌出量 预测 

分 类 号：TD712.5[矿业工程—矿井通风与安全]