MC-最优变权组合模型在建筑物变形中的应用  

Application of the MC-optimal variable weight combination model for buildings deformation

作　　者：唐诗华[1,2] 王江波 王凯 李飞达 TANG Shi-hua;WANG Jiang-bo;WANG Kai;LI Fei-da(a.Guangxi Key Laboratory of Spatial Information and Geomatic;b.College of Geomatics and Geoinformation,Guilin University of Technology,Guilin 541004,Chin)

机构地区：[1]桂林理工大学广西空间信息与测绘重点实验室,广西桂林541004 [2]桂林理工大学测绘地理信息学院,广西桂林541004

出　　处：《桂林理工大学学报》2018年第2期295-300,共6页Journal of Guilin University of Technology

基　　金：国家自然科学基金项目(41461089);广西自然科学基金项目(2014GXNSFAA118288);广西空间信息与测绘重点实验室项目(15-140-07-05;16-380-25-13)

摘　　要：提出一种基于马尔科夫链(MC)误差修正的最优变权组合预测模型。该模型用马尔科夫链误差修正理论分别对GM(1,1)模型和线性回归模型的预测值作进一步的修正,充分发掘残差序列隐含信息,再建立最优变权组合预测模型,给出了MC-最优变权组合预测模型算法的实现流程。通过算例验证表明,该模型继承各单一模型的优点,并克服其模型缺陷,同时具备了马尔科夫链的误差修正特性,在一定程度上可以确保较优的局部预测数据和较好的全局预测精度,在建筑物变形预测中具有应用价值。In the development of current forecasting models,an optimal variable weight combination forecasting model based on Markov chain（ MC） error correction is proposed. The GM（ 1,1） model and the linear regression model are further modified by the Markov chain error correction theory. The implicit information in the residual series can be fully exploited and the optimal combination forecasting model of variable weight is established. The realization process of the algorithm of MC-optimal variable weight combination forecasting model is presented.Numerical examples show that the model inherits the advantages of the single model and has the error correction property of the Markov chain. It can ensure better local prediction data and better global prediction to a certain extent. The results show that the model has certain practical value in the prediction of buildings deformation.

关 键 词：GM(1 1)模型 线性回归模型 马尔科夫链误差修正 最优变权组合模型 建筑物变形预测 

分 类 号：P207[天文地球—测绘科学与技术]