基于PCA-KFDA的煤矿火区启封复燃预测  

Prediction of Resurgence of Fire Zone after Unsealing Based on PCA-KFDA

作　　者：田少波[1] 卢国斌[1] 

机构地区：[1]辽宁工程技术大学矿业学院,阜新123000

出　　处：《世界科技研究与发展》2016年第5期965-968,1005,共5页World Sci-Tech R&D

基　　金：国家自然科学基金(51374121)资助

摘　　要：为准确、快速地对煤矿火区启封是否复燃进行预测,针对煤矿火区启封复燃参数之间信息重叠和非线性的问题,提出一种PCA-KFDA煤矿火区启封复燃预测方法。结合我国典型矿井火区启封实例,选取煤自燃倾向、φ(C_2H_2)、φ(CO)、φ(CO_2)、φ(C_2H_4)和φ(O_2)(φ表示气体在空气中的体积分数)作为煤矿火区启封复燃预测的影响因素,建立煤矿火区启封复燃预测的PCA-KFDA模型。利用该模型进行实际预测,预测结果与实际情况完全吻合。而利用Fisher判别模型、神经网络模型、Logistic模型进行预测均出现不同程度的误差,验证了本文提出模型的有效性和准确性。In order to predict whether the fire zone will resurgence after unsealing accurately and quickly,a PCA-KFDA method is proposed,which can overcome the problem of information overlap and nonlinear. Combined with the examples of typical mine the reopening of fire area in China,coal spontaneous combustion tendency、φ（ C2H2） 、φ（ CO） 、φ（ CO2） 、φ（ C2H4） and φ（ O2）（ where φ represents the volume fraction of the gas in the air） are selected as the influencing factors of the coal mine in fire zone unsealing resurgence prediction,and a PCA-KFDA model is established using the obtained influencing factors. Then,the model is used to predict the actual situation,and the predicted results are in agreement with the actual situation. But the results of Fisher discriminant model,neural network model,Logistic model all have varying degrees of errors,which verifies the validity and accuracy of the proposed model.

关 键 词：煤矿火区 火区启封 主成分分析 核FISHER判别分析 

分 类 号：TD752.2[矿业工程—矿井通风与安全]