摘要
紧邻露天矿边坡的地表受采剥过程中多种因素耦合的影响,出现移动变形,从而威胁到了毗邻露天矿的工业及民用建筑物的安全使用,为了预测其变形值,通过建立增加动量项的自适应BP神经网络,借助已有样本数据,对网络进行训练、测试,经与目标值对比,其精确度较高。将训练好的网络应用于观测网络的测点预测,得到了其2013年的沉降变形值。该方法对多因素耦合作用的地表沉陷预计具有重要意义。
Adjacent to the slope surface which is affected by many coupling factors in the process of stripping appear deformation,which threatens the safe using of adjacent pit industrial and buildings,to predict the surface deformation. Establish an adaptive and momentum BP neural network, through the sample data,the network is trained and tested. The results show that has high accuracy compared with the target value.The trained network is applied in the measuring point of observation network measurement, the settlement and deformation of the 2013 were obtained. The method which multi factors coupling to surface subsidence has important significance.
出处
《煤炭技术》
CAS
北大核心
2015年第8期100-102,共3页
Coal Technology
关键词
露天矿边坡
多因素耦合
BP神经网络
地表变形
预测
open pit slope
multi-factor coupling
BP neural network
ground settlement
prediction