STUDY ON SPATIAL MULTIPOINT RESIDUAL MODEL BASED ON GREY PREDICTION
Li Xiaolei 1,3) ; Liu Rui 2) ; Tian Yongrui 3) ;and Xie Bingke 1)
1)School of Civil Engineering and Architecture, Chongqing University of Science and Technology, Chongqing 401331 2)Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 1001013)College of Geology Engineering and Geomatics, Chang’an University, Xi’an 710054
Abstract：On the basis of the analysis of the shortage of Grey singlepoint prediction model, by considering the spatial correlation between the deformation observation points, the residual modification is applied in the spatial multipoint model and thus the spatial multipoint residual model is developed.By taking the subsidence monitoring data of Xiaolangdi multipurpose dam as an example, the results shows that the total absolute values of relative error(5 cycles) of spatial multipoint residual model is 122%, better than 179% of spatial multipoint model and 284% of GM(1,1) model. Meanwhile, this model has the best fitting accuracy with the relative error less than 20% in shortcycle fitting.
