Kalman-GM(1,1)组合模型在基坑围护墙顶沉降预测中的应用

针对基坑围护墙顶沉降监测数据受外界随机噪声干扰较大的问题,提出利用Kalman-GM(1,1)组合模型来进行变形分析和预测。即先用Kalman滤波模型对观测数据进行去噪处理,再建立基于滤波数据的GM(1,1)模型,进行基坑墙顶沉降预测。工程实例应用表明,该组合模型有效减弱了随机噪声干扰,其预测精度和可靠性高于单一GM(1,1)模型,更适用于基坑墙顶沉降预测。

Application of Kalman-GM ( 1,1) Combined Model in Settlement Prediction of the Top of the Enclosure Wall of the Foundation Pit

Aiming at the problem that settlement monitoring data of the top of the enclosure wall of the foundation pit is interfered with random noise,this paper proposes using Kalman-GM(1,1)combination model to analyze and predict deformation.Firstly,the Kalman filter model is used to denoise the observed data,and then the GM(1,1)model is used to predict the settlement data based on the filtered data.The application example shows that the combined model effectively reduces the random noise interference,and its prediction accuracy and reliability are higher than the single GM(1,1)model,which is more suitable for settlement prediction of the top of the enclosure wall of the foundation pit.