MGM(1,n)模型在粉喷桩路基沉降预测中的应用  

Application of multivariate gray model in prediction of deep-mixing pile compound roadbed settlement

作　　者：陈斌[1] 陈晓东[1] 张仪萍[2] 

机构地区：[1]浙江水利水电专科学校水利系,浙江杭州310018 [2]浙江大学建筑工程学院,浙江杭州310027

出　　处：《辽宁工程技术大学学报（自然科学版）》2007年第3期378-380,共3页Journal of Liaoning Technical University (Natural Science)

基　　金：国家自然科学基金资助项目(50278089);浙江省自然科学基金资助项目(501163)

摘　　要：提出了一种预测公路复合地基工后沉降的新方法。首先,通过路基监测项目的灰色关联分析,优选影响路基沉降的相关预测变量,如道肩水平位移、孔隙水压力消散值及桩身应变等,在此基础上建立灰色多变量MGM(1,n)模型并给出相应的沉降预测方程。而后考虑监测数据序列的非等时间间隔影响,改进、完善此模型,使之能够对粉喷桩复合路基的沉降量进行准确预测。文中并给出应用实例,表明应用该法进行沉降预测,可以更全面地利用工程监测数据,更有效地预测路基工后沉降量。A new method to forecast after-construction settlement of compound roadbed is put torward, l-lrstly, by means of roadbed monitoring-item＇s gray associate analysis, optimal selection of variables such as shoulder horizontal displacement, pore water pressure slaking valuation, pile body strain and so on, can be made. Secondly, using the variables selected in step 1, the MGM（1,n） model and its corresponding forecast equation can be established. Thirdly, considering the effect of monitoring-item＇s unequal time-interval, the model established can be consummated and improved. Finally, by using the established model, the settlement of soft roadbed can be accurately forecast. As a case study, a practical example is given, which indicates the method can make use of monitoring data more fully, and can forecast after-construction settlement more effectively.

关 键 词：路基沉降 预测 非等时距:MGM(1 n)模型 

分 类 号：U416.1[交通运输工程—道路与铁道工程]