残差修正的GM(1,1)模型在滑坡位移预测中的应用

Application of Residual Modified GM(1, 1) Model in Landslide Displacement Prediction

根据滑坡监测研究现状,利用Matlab编程以及新滩滑坡监测点A3和B3的位移监测资料,建立了GM(1,1)预测模型和一阶残差修正的GM(1,1)预测模型;并对两种灰色预测模型的变形预测结果进行了比较。结果表明,经过残差修正的GM(1,1)模型的预测精度明显高于传统GM(1,1)模型的预测精度。

According to the research status of landslide displacement monitoring, we used the Matlab programming and the displacement monitoring data of Xintan landslide monitoring points A3 and B3 to build GM(1,1) prediction model and the rst-order residual modified GM(1,1) model, and compared the deformation prediction results of two gray prediction models. The experimental result shows that the prediction accuracy of the residual modi ed GM(1,1) model is significantly higher than that of the traditional GM(1,1) model.