基于多分类支持向量机的隧道挤压大变形预测

Prediction of tunnel squeezing based on multiclass Support Vector Machines

介绍了隧道挤压大变形的危害,提出了一种基于四个参数即直径(D),埋深(H),支护刚度(K)和岩石隧道开挖质量指数(Q)的多分类支持向量机(SVM)方法,对隧道挤压大变形的预测进行了研究,解决了隧道变形多分类预测问题,提高了预测精度。

The harm of tunnel squeezing is introduced.The multiclass Support Vector Machine(SVM)method based on four factors,diameter(D),depth(H),supporting stiffness(K)and rock tunneling quality index(Q),is proposed.The prediction of tunnel squeezing is studied.The problem of multiclass prediction of tunnel squeezing is solved,and the prediction accuracy is improved.