开采引起地表沉降的神经网络预测研究  

Prediction of Surface Subsidence in Underground Mining with Neural Network

作　　者：王建华[1,2] 康志强[1] 吕广忠[1] 王福生[1] 杨国华[3] 

机构地区：[1]河北理工大学 [2]河北省武安市冶金矿山管理局 [3]河北省矾山磷矿

出　　处：《现代矿业》2009年第5期66-69,共4页Modern Mining

摘　　要：介绍了矿山地下开采引起地表下沉对地表建筑设施造成了的损害,分析了地表沉陷受地质条件和采矿条件等诸多因素的影响,且因素之间存在非线性关系,难以用数学模型加以描述,因而求解困难。研究利用神经网络系统对地表沉陷问题进行预测,通过试验采集的数据对神经网络进行了训练和检测,获得了比较满意的结果,预测值和期望输出值之间存在着极小的误差,从而证明了对地表沉陷预测的可行性和实用性。Underground mining causes surface subsiding, resulting in damage of surface building Surface subsidence is limited by many factors such as geologic conditions and mining conditions. There are nonlinear relations between the factors and it is difficult to describe it by a mathematical model, therefore, there are difficulties in solving some problems. The surface subsidence is predicted by the neural network system. The system is trained and tested by the data collected in tests. It has achieved satisfacto- ry results, and there are very small errors between the predicted and the expected values. The results prove that it is feasible to use neural network to predict the surface subsidence.

关 键 词：B^P神经网络 地表沉陷预测 矿山开采 

分 类 号：TD325.2[矿业工程—矿井建设]