费歇判别分析方法在采空区塌陷预测中的应用  

Application of Fisher Discriminant Method in Goaf Collapse Prediction

作　　者：白云飞[1] 邓建[1] 董陇军[1] 李响[1] 

机构地区：[1]中南大学资源与安全工程学院,湖南长沙410083

出　　处：《矿业研究与开发》2008年第5期74-76,共3页Mining Research and Development

基　　金：国家自然科学基金项目资助(50574098)

摘　　要：将费歇判别分析法应用到采空区塌陷的预测中,选取覆盖层类型、覆盖层厚度、地质构造复杂程度、煤层倾角、采空区体积率、采空区距地表的垂直深度、采空区空间叠置层数等7个影响因素,建立了采空区塌陷预测的费歇判别模型。利用北京西山某地区采空区17组实测数据作为训练样本逐一回判,其误判率为零,另外7组预测样本的判别分析结果与实测结果也完全吻合。研究结果表明:采空区塌陷预测的费歇判别模型判别性能良好,判别准确度高,且计算过程简单,可操作性强,是采空区塌陷预测中的一种非常有效的方法。Based on the Fisher Discriminant theory, the type and thickness of overburden layer, the complex degree of geologic structure, the inclination angle of coal seam, the volume rate of goaf, the vertical depth from goaf to surface and space superposition layer of goaf are used as the distinction factors of the Fisher discriminant model for goaf collapse prediction. The 17 groups of measured data of one goal in Xishan, Beijing, were used as the training sample, and the Backexamining of the training sample was carried out one by one using the model, and the error rate of discrimination is zero, the discriminating results of other 7 groups of forecast sample were agreed well with the actual results. The studied results indicated the Fisher Discriminant method of goaf collapse prediction has good performance and high discriminating accuracy,also has the advantages of simple computation process and good operation. Therefore, it is one extremely effective method in forecasting goaf collapse.

关 键 词：采空区塌陷 费歇判别法 数值预测 

分 类 号：TD327[矿业工程—矿井建设]