期刊论文

中文

长沙理工大学学报(自然科学版)

1672-9331

2012

9

4

32-36

10.3969/j.issn.1672-9331.2012.04.006

全国教育科学‘十一五’规划资助项目（FFB09573）；

本校为其他机构

为了在贫数据条件下准确预报中长期沉降值，采用线性回归方程和指数方程的组合方式，通过适当配置模型的某些参数来获得新的生成序列函数模型．结合工程实际算例进行预测，并和实际观测数据比较，取得了较好的效果，验证了灰色线性回归组合模型应用于沉降监测预报的可行性．通过其残差与回归分析模型和灰色预测GM（1，1）模型残差相比较，证明该模型在沉降预报中优于回归分析模型和灰色预测GM（1，1）模型．

In order to predict long-term settlement accurately in the poor condition of data, this article adopted linear regression equation and exponential equation combination model. Combination with the engineering practice example forecast and actual observation data comparison and good xesult have been achieved. It is verified that linear regression com- bined model can be applied to forecast the feasibility of subsidence monitoring. Through comparing residuals and regression analysis with grey prediction GM （1,1） model residual phase, this paper ...