改进GM(1,1)在高铁隧道沉降变形预测中的对比应用

针对传统GM(1,1)模型在高铁隧道沉降变形分析与预测中精度不理想状况,本文在传统GM(1,1)模型基础上,建立自适应GM(1,1)模型与残差修正GM(1,1)模型并讨论两种改进模型各自优点。利用传统GM(1,1)模型、自适应GM(1,1)模型以及残差修正GM(1,1)模型对某高铁隧道监测点作沉降分析与预测。通过对比,得出自适应GM(1,1)模型与残差修正GM(1,1)模型对原模型的预测曲线相关性和预测精度有一定程度提高;残差修正GM(1,1)模型对于沉降曲线波动较大处仍有较好的拟合与预测效果,其预测效果优于自适应GM(1,1)模型。

Comparison and Application of Improved GM(1,1) in High-speed Railway Tunnel Settlement Deformation Prediction

Aiming at the situation of the precision of traditional GM (1,1) in high-speed railway tunnel settlement deformation analysis and prediction is not ideal .This paper which is based on traditional GM (1,1) model has estab-lished self-adaptive GM(1,1) model and residual error correction GM (1,1) model and discussed their respective ad-vantages.Using traditional GM (1,1) model, self-adaptive GM(1,1) model and residual error correction GM (1,1) model to analyze and predict a High-speed Rail tunnel monitoring points settlement deformation .Through comparing and analyzing, it is concluded that self-adaptive GM (1,1) model and residual error GM (1,1) model improve the predic-tion precision of original model and the correlation of prediction curve in a certain extent;residual error correction GM (1, 1 ) model has a better fitting and prediction effect for the settlement curve with bigger fluctuations , its prediction effect is superior to the self-adaptive GM (1, 1) model.