硫化矿石自燃倾向性等级分类的Fisher判别分析法  

Fisher discriminant analysis to the classification of spontaneous combustion tendency grade of sulphide ores

作　　者：胡汉华[1,2] 刘征[1,2] 李孜军[1,2] 崔田田[1,2] 

机构地区：[1]中南大学资源与安全工程学院,湖南长沙410083 [2]中南大学国家金属矿安全科学技术研究中心,湖南长沙410083

出　　处：《煤炭学报》2010年第10期1674-1679,共6页Journal of China Coal Society

摘　　要：将Fisher判别分析法(FDA)应用于硫化矿石自燃倾向性等级分类问题中,选用主要矿物及含量、吸氧速度常数平均值、自热点、自燃点4项指标作为判别因子,建立硫化矿石自燃倾向性等级分类的FDA模型。选取新桥硫铁矿13个矿样的实验数据作为学习样本进行训练,建立相应的线性判别函数,并用训练后的模型对天马山高硫金矿的-55m中段以上矿石自燃倾向性进行判别。结果表明:Fisher判别分析模型分类性能良好,对硫化矿石自然发火潜势的划分结果统计意义明显,适用性强,说明该方法在硫化矿石自燃倾向性判别与分类中有较好的应用前景。Based on the principle of Fisher discriminant analysis,a classification model of spontaneous combustion tendency of sulphide ores was established,including four indexes reflecting the spontaneous combustion tendency of sulphide ores：compositions and contents of mineral,velocity constant of oxygen absorption,spontaneous heating point and spontaneous combustion point.Linear discriminant functions were obtained through training thirteen experimental data of ore samples from Xinqiao pyrite.Based on this,spontaneous combustion tendency of sulphide ores from Tianmashan high sulfur gold were discriminated.The results show that the classification model of Fisher discriminant analysis has excellent performance and a good applicability,and statistic significance of classifying is explicit,from which we can see that it has a good prospect for the identification and classification of spontaneous combustion tendency of sulphide ores.

关 键 词：硫化矿石 自燃倾向性 等级 Fisher判别分析法 分类 

分 类 号：TD752[矿业工程—矿井通风与安全]