基于支持向量机的露天转地下开采边坡变形模型  

Slope deformation model of metal mines transferred underground mining from open-pit based on support vector machines

作　　者：李长洪[1,2] 王云飞[1,2] 蔡美峰[1,2] 苗胜军[1,2] 范丽萍[1,2] 

机构地区：[1]北京科技大学金属矿山高效开采与安全教育部重点实验室,北京100083 [2]北京科技大学土木与环境工程学院,北京100083

出　　处：《北京科技大学学报》2009年第8期945-950,共6页Journal of University of Science and Technology Beijing

基　　金：国家科技支撑计划资助项目(No.2006BAB02A17)

摘　　要：提出了一种基于支持向量机的露天转地下开采边坡变形模型,有效表达了地下开采扰动引起露天矿边坡变形的非线性变化关系.采用RBF核函数学习现场监测数据,利用交叉验证选择模型参数,通过学习捕捉支持向量,建立模型预测未来变化趋势.将该模型应用于露天转地下开采的杏山铁矿.结果表明,支持向量机对学习样本的拟合精度极高,其预测精度也很高.采用捕捉的支持向量进行预测,便捷快速且有较强泛化能力.A slope deformation model of metal mines transferred underground mining from open-pit based on support vector machines was presented. The model can effectively express the non-linear variation of metal mine open-pit slope deformation caused by underground mining disturbance. In the model the RBF kernel function was utilized to train on-site monitoring data, the cross-valida- tion was employed to choose model parameters, support vectors were achieved with training samples, and then the future deformation was predicted. The model was applied to Xingshan Iron Ore transferred underground mining from open-pit. The results show that the regression value of learning samples is extremely precise and the predicted deformation has a higher precision based on support vector machines. The application of the model, which predicts the deformation with the achieved support vectors, is convenient and it bears a stronger generalization ability.

关 键 词：岩土工程 地下开采 支持向量机 边坡变形 

分 类 号：TD854.6[矿业工程—金属矿开采]