小波去噪的灰色动态模型在变形预测中的应用

针对GM(1,1)模型对随机波动性较大的数据拟合较差、预测精度低的缺点,提出了基于小波去噪的灰色动态模型。首先运用小波滤波消除数据噪声,使数据更具规律性;再利用灰色动态模型预测变形;最后对高层建筑物沉降监测数据的预测值与实测值进行对比分析。结果表明,该模型的预测误差较小、精度较高,适合在变形预测中应用。

Application of Gray Dynamic Mode in Deformation Prediction Based on Wavelet Denoising

GM(1,1) model of random volatile data fit is poor.Because of the shortcomings of low prediction accuracy,the gray dynamic model based on wavelet denoising was put forward in this paper.Firstly,it used wavelet filtering to eliminate data noise,then re-used gray dynamic model to predict deformation.Finally,through the high-rise building subsidence monitoring data predicted and measured values of the comparative analysis showed that the model prediction error was small,high accuracy,suitable for application in deformation prediction.