摘要
在隧道工程中,位移变形监测是一个重要环节,然而监测数据往往呈现出较为复杂的非线性特征。以铜黄高速公路大田双连拱隧道施工为背景,在支持向量机算法的基础上,采用变量轮换法对其参数进行优化处理,从而呈现了监测数据复杂非线性特征并建立隧道位移监测时间序列非线性模型。利用此模型对未来的位移变形做出精确的预报,科学地指导现场监测和施工建设。
During the process of tunnel construction,monitoring displacement of tunnel plays an important role, but monitoring data are always characterized by complex nonlinear character.Under the background of DATIAN doub- le-arch tunnel in construction of the Tonghuang Highway,cyclic variable method is used to optimize the parameters of SVM based on SVM algorithm in order to present the nonlinear relation between the data,meanwhile the nonlinear model for displacement is established.That is the model is introduced to forecast anficipative deformation and guide monitoring and construction in field scientifically.
出处
《地下空间与工程学报》
CSCD
2007年第z1期1346-1348,1353,共4页
Chinese Journal of Underground Space and Engineering
关键词
隧道工程
支持向量机
变量轮换法
位移预报
tunnel engineering
support vector machine
cyclic variable method
displacement prediction
