Title :
Application of GM(1,1) Optimized Model in Prediction of Landslide
Author :
Jin, Xiao-Guang ; Zeng, Jie ; Liu, Xin-Rong
Author_Institution :
Chongqing Univ., Chongqing
Abstract :
Grey system theory is a new cross course founded in 1982 by Prof. Deng Ju-long, a Chinese famous scholar. Grey model of the theory was applied extensively and gained a series of significant achievement both in natural science and social science. It is mainly applied in mid-long-term forecasting of slope or landslide and has high accuracy in geologic catastrophe study. However, when applied in short-impending slip forecasting, it produces relatively poor accuracy, even not suit the study. There is still much room for its improvement. Actual counting examples of slide deformation forecasting indicated that optimized model GM(1,1) based on ground value of optimized grey model had an ability to optimize modeling results; it can be applied with relative high modeling and forecasting accuracy not only in mid-long-term forecasting of slope deformation, but also in short- impending slip forecasting of landslide. In this paper, correctness and extensive application of the optimized model GM (1,1) were tested by analysis of the contrast between original linear model GM(1,1) and non-linear model Verhulst. The study will not only enrich research methods of landslide deformation forecasting, but is also of momentous theoretical and practical significance.
Keywords :
geomorphology; grey systems; matrix algebra; GM(1,1) optimized model; Grey model; Grey system theory; geologic catastrophe; landslide deformation forecasting; slope deformation; Civil engineering; Deformable models; Differential equations; Educational institutions; Geology; Optimization methods; Predictive models; Terrain factors; Testing; Uncertain systems; 1); application; deformation; landslide; modeling and forecasting; optimized model GM (1;
Conference_Titel :
Natural Computation, 2007. ICNC 2007. Third International Conference on
Conference_Location :
Haikou
Print_ISBN :
978-0-7695-2875-5
DOI :
10.1109/ICNC.2007.254