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.
出处
《煤炭学报》
EI
CAS
CSCD
北大核心
2010年第10期1674-1679,共6页
Journal of China Coal Society
关键词
硫化矿石
自燃倾向性
等级
Fisher判别分析法
分类
sulphide ores
spontaneous combustion tendency
grade
Fisher discriminant analysis
classification
