基于综合优化GM(1,1)的形变预测模型  

Deformation Forecasting Model Based on Synthetic and Optimized GM(1,1)

作　　者：李克昭[1,2] 李志伟[1] 孟福军[3] 丁安民[1] 

机构地区：[1]河南理工大学测绘与国土信息工程学院,焦作市454003 [2]北斗导航应用技术协同创新中心,郑州市450001 [3]塔里木大学水利与建筑工程学院,阿拉尔市843300

出　　处：《大地测量与地球动力学》2016年第2期120-123,128,共5页Journal of Geodesy and Geodynamics

基　　金：国家自然科学基金(41202245,41272373);河南理工大学骨干教师基金(72105/090)

摘　　要：通过对传统GM(1,1)模型的初始条件进行优化,利用复化的Simpson数值积分公式构造背景值,以及对残差序列进行修正,建立优化的GM(1,1)模型,并给出优化模型的精度评价方法和标准。以小浪底大坝的变形实测数据为例,进行基于传统GM(1,1)预测模型和优化GM(1,1)预测模型的计算比较。结果表明,基于优化的GM(1,1)预测模型精度更高。Based on the optimization to the initial conditions of the traditional model GM（1,1）,the use of background value in construction of compound Simpson numerical integral formula,and the modification on the residual sequence,the optimized GM（1,1）model is proposed in this paper.Finally,on the basis of measured data of Xiaolangdi dam＇s deformation,the calculation results based on traditional GM（1,1）forecasting model with based on optimized GM（1,1）forecasting model are compared.The results show that the based on optimized GM（1,1）prediction model of the accuracy is better than the based on traditional GM（1,1）forecasting model.

关 键 词：变形灾害预测 GM(1 1)模型 复化Simpson公式 残差修正 

分 类 号：P258[天文地球—测绘科学与技术]