改进的GM(1,1)模型在大坝变形预测中的应用

Application of improved GM(1,1)model to dam deformation forecasting

针对大坝安全监测数据存在贫信息、小样本的问题,通过对原始数据、初始值以及背景值进行优化改进传统的GM(1,1)模型。以某大坝实测径向水平位移数据为例,分别用改进前后的模型进行预测,并与实测值进行对比。验证改进的GM(1,1)模型的优越性与有效性,相对于传统GM(1,1)模型,其预测精度更高。

For dam safety monitoring data ,poor information and small sample problem are unavoidable .In this paper ,the traditional GM (1 ,1) model is improved by optimizing the original data ,the initial value and the background value .Taking the radial horizontal displacement samples of a dam ,the two models are adopted to predict the displacement values respectively ,and compared with the measured values .Compared with the traditional GM(1 ,1) model ,the improved GM(1 ,1) model is effective and practical ,of which the prediction accuracy is higher .