煤层注水难易程度的Fisher判别分析模型及应用  

Fisher discriminant analysis model and application of classifying difficulty degree of water infusion for coal seam

作　　者：袁志刚[1] 王宏图[1,2] 胡国忠[3] 刘年平[1] 范晓刚[1] 

机构地区：[1]重庆大学西南资源开发及环境灾害控制工程教育部重点实验室,重庆400030 [2]重庆大学复杂煤气层瓦斯抽采国家地方联合工程实验室,重庆400030 [3]中国矿业大学矿业工程学院,江苏徐州221116

出　　处：《煤炭学报》2011年第4期638-642,共5页Journal of China Coal Society

基　　金：国家自然科学基金资助项目(50774106);国家自然科学创新群体基金资助项目(50921063);川煤集团科技资助项目(2009-08)

摘　　要：基于Fisher判别法理论和影响煤层注水难易程度的重要因素,选用煤层的埋藏深度、裂隙的发育程度、孔隙率、湿润边角、饱和水分增值、坚固性系数6项指标作为判别因子,以15组煤层注水的实测数据作为学习样本进行训练,建立相应的Fisher判别函数对15组实测数据用回代估计方法逐一进行检验,正确率为93.3%。将建立的模型应用于3个工程实例,与实际情况吻合良好,计算结果与传统的模糊聚类分析法、神经网络法所得到的结果一致或更好。Based on Fisher discriminant theory and main factors influencing difficulty degree of water infusion for coal seam,six major difficulty degree of water infusion factors for coal seam such as buried depth,cranny viability,porosity,wet rim angle,saturation moisture increment,firm coefficient of coal were regarded as the discriminant factors of the Fisher discriminant analysis model;Discriminant functions were obtained through training of fifteen sets of in-situ data,each of the fifteen sets of samples was tested by using resubstitution method according to the Fisher discriminant method,and the correct rate was equal to 93.3% after the Fisher discriminant analysis was trained.Moreover,the proposed method was applied and tested in three actual water infusion cases,both the validity and the feasibility were approved.The result is the same or better than the results with the Fuzzy classification analysis method and neural network method.

关 键 词：煤层注水 注水难易程度 Fisher判别法 

分 类 号：X913[环境科学与工程—安全科学]