Logistic及Verhulst预测模型在路基沉降预测中的拟合研究  

Research on the Fitting of Logistic Model and Verhulst Model for Subgrade Settlement Prediction

作　　者：胡晓阳[1] 王连俊[1] 张光宗[1] 

机构地区：[1]北京交通大学土木建筑工程学院,北京100044

出　　处：《铁道标准设计》2013年第9期1-4,共4页Railway Standard Design

基　　金：铁道部科技研究开发计划项目(2008G032-A)

摘　　要：高速铁路路基沉降是一个复杂的系统过程,涉及不同时期、不同工况,选用单一的预测模型不足以反映路基整体的变形规律。Logistic模型及Verhulst模型的曲线均为S形,可以反映全过程的地基沉降与时间的关系。针对路基填筑期和降水期两个时段利用MATLAB软件编制程序计算模型参数,得到Logistic模型和灰色Verhulst模型,进而求得两者的最优加权组合模型。对比结果表明:在路基填筑期由两者的最优加权组合模型预测得到的结果与实测结果较吻合,拟合精度高于单一模型;而在降水期Verhulst模型的预测值更接近实测值,平均相对误差约为1.75%,拟合精度较高。Subgrade settlement of high-speed railway is a complex systematic process that involves different periods and construction conditions, so only choosing a single model cannot represent the integrated deformation rule of subgrade. Both the Logistic model and Verhulst model can be described by S-shaped curve, and can reflect the settlement-time relation of subgrade through the whole process. Focusing on the two periods of subgrade filling period and precipitation period respectively, by means of the MATLAB software to program and calculate the model parameters, this paper obtained the Logistic model and Verhulst model, and then obtained the optimal weighted combination model of the both. Comparative calculation results show that the optimal weighted combination model of the both can fit well with the measured date, and the fitting precision is better than that of each single model in subgrade filling period; while in the precipitation period there is higher predictive precision by only using the Verhulst model that is more close to the measured data, with an average relative error of about 1.75%.

关 键 词：高速铁路 预测模型 MATLAB程序 路基沉降 拟合 

分 类 号：U213.1[交通运输工程—道路与铁道工程]