基于OWGA算子的隧道变形最优组合预测  

Optimal Combination Prediction of Tunnel Deformation Based on OWGA Operators

作　　者：刘绍堂[1] 刘文锴[1] 张慧锋[1] 潘洁晨[1] 

机构地区：[1]河南工程学院,郑州451191

出　　处：《铁道工程学报》2014年第7期84-88,共5页Journal of Railway Engineering Society

基　　金：河南省重点科技攻关计划(122102210421);河南省教育厅科技攻关计划(2010A420001)

摘　　要：研究目的:根据隧道变形数据预测隧道的变形发展情况,结合具体工程实践,采用最新非线性预测理论,在文献综述的基础上,结合对数灰关联度与OWGA算子,把预测值和观测值的对数序列之间的灰关联度设定为目标函数,有序加权集成做为诱导变量值的预测精度,提出一种新最优隧道变形预测模型。采用MATLAB编程分别实现灰色预测、指数平滑预测、自回归移动平均预测,并利用基于对数灰关联度的有序加权几何平均组合预测模型,在对数灰关联度最大的意义下求出组合模型的最优权重,通过隧道变形预测的实例分析,说明新模型的有效性和合理性。研究结论:(1)基于OWGA算子的隧道变形预测,在预测误差平方和、均方误差、平均绝对误差、平均绝对百分比误差、均方百分比误差等预测有效性精度指标下,比灰色预测,平滑预测,ARMA预测等预测方法的预测精度都高,误差最小;(2)通过工程实例检验,该预测方法的预测结果不但合理、有效,而且是优性预测;(3)本研究成果可应用于隧道安全监测和变形预测领域。Research purposes：In order to predict tunnel deformation developments,this paper proposed the optimal combination prediction model of tunnel deformation,based on literature review and the number of gray relational degree OWGA operators,using the latest nonlinear predictive theory,combined with specific engineering practice.Taking the logarithm of the gray relational degree between predictive value sequence and the observational value sequence as the objective function,integrated ordered geometric weight based on prediction accuracy.The analyses using the MATLAB in Grey,exponential smoothing,ARMA and the new model are to test the effectiveness and the rationality of the new model.Research conclusions：Compared to grey prediction,smoothing prediction,ARMA prediction,the prediction accuracy of optimal combination prediction of tunnel deformation based on the number of gray relational degree OWGA operators is higher with minimum error.Its forcasting result is not only reasonable and effective,but also is optimal.This research can be applied to tunnel safety monitoring and deformation prediction field.

关 键 词：隧道变形预测 隧道拱顶下沉 OWGA算子 对数灰关联度 最优组合预测 

分 类 号：U25[交通运输工程—道路与铁道工程]