半参数改进灰色模型在滑坡变形预测中的应用

Application of improved semiparametric gray model in landslide deformation prediction

滑坡灰色模型的模型误差主要来自降雨量、温度等外界影响因子,传统的半参数灰色模型没有考虑这些对滑坡变形影响较大的外界因子,而把相邻时刻的模型误差当作是不变的,预测精度较低。针对这一问题,该文提出了将这些影响因子当作非参数变量引入模型,通过改进正规矩阵来建立半参数改进灰色模型,可以得到更加准确的模型误差,并且能够将其补偿到观测序列中,使预测结果更加准确。计算结果表明,本文所述模型在观测序列的拟合和预测中均有较好的结果,能够充分地利用在滑坡中采集到的各种信息,并且达到更优的结果。

The model error of landslide gray model mainly comes from external influence factors such as rainfall,temperature and so on,the traditional semiparametric gray models didn’t consider these influence factors,it regarded the model error of the adjacent time as constant,so the forecast accuracy was lower.In view of this problem,this paper tried to regard these influence factors as nonparametric variables,then introduced into the model,establishing a semiparametric improved gray model by improving the normal matrix,a more accurate model error could be obtained,and the model error could compensate to the observation sequence,then we could get more accurate prediction result.The calculation results showed that the model described in this paper had better results in the fitting and prediction of the observation sequence,the model described in this paper could make full use of all kinds of information collected in landslides and achieve better results.