矿区采空塌陷危险性预测的Bayes判别分析法

The Bayes Discriminant Method for Forcasting the Stability of Underground Goaf

提出了矿区采空塌陷危险性预测的Bayes判别分析方法.首先分析了目前因采空区塌陷导致灾害事故发生的危害程度,随后介绍了Bayes判别分析方法的原理、判别准则和检验方法,在此基础上考虑采空区塌陷问题的具体特点和影响因素,选取北京西山某矿区的典型塌陷情况作为案例,对Bayes方法的运用进行了具体说明.分别选取了覆盖层类型、覆盖层厚度、地质构造复杂程度、矿层倾角、采空区体积率、采空区距地表的垂深和采空区空间叠置层数作为判别指标,以历史上17个典型塌陷资料作为学习样本进行训练,建立采空塌陷危险性预测的Bayes判别分析模型,并利用回代估计法对该模型进行检验.研究结果显示,训练后的模型判别结果完全符合实际情况.将该模型运用于7个预测样本的判别中,判别结果也和实际情况一致,说明Bayes判别分析模型具有良好的判别能力.

The Bayes discriminant analysis method for forecasting the stability of underground goaf is presented in this paper.The disaster degree induced by underground goaf collapse is summarized firstly,and then the principle,theory and testing method of Bayes discriminant analysis method introduced.Based on the historical collapse information of Beijing Xishan Mine and by considering the characteristics of underground goaf,the specific process of Bayes discriminant analysis method is demonstrated in the paper.Seve...