灰色GM(1,1)与Kalman滤波模型用于变形预测的比较分析

Comparative Analysis of Applications of Grey System GM(1,1) Model and Kalman Filter Model in Building Settlement Deformation

通过运用灰色GM(1,1)和Kalman滤波模型分别对某高层建筑物的沉降变形趋势进行分析,得出灰色GM(1,1)适用于短期且变形趋势呈线性变化的变形分析与预测,而Kalman滤波模型不仅适用于短期预测,对于长周期预测也有较高的精度。因此,在观测周期较短时,灰色GM(1,1)和Kalman滤波模型对线性变化的高层建筑物都有较高的预测精度,但是对于较长周期的观测,Kalman滤波模型预测的精度和可靠性要高于灰色GM(1,1)模型。

In this paper,we used Grey system GM(1,1) model and Kalman filter model to predict the trend of a High-rise building's deformation at first.The results show that Grey system GM(1,1) model is applied to predicting and analyzing the short-period deformation.And the Kalman filter model not only is applied to predicting and analyzing the short-period deformation,but also it has a high accuracy in a long-period deformation prediction.