Displacement prediction of step-like landslide based on the response of inducing factors and support vector machine

作者：Zhou, Chao; Yin, Kunlong; Cao, Ying; Huang, Faming

来源：Chinese Journal of Rock Mechanics and Engineering, 2015, 34: 4132-4139.

DOI：10.13722/j.cnki.jrme.2014.0290

• Deformation Characteristics
• Displacement prediction
• Displacement time series
• Geological conditions
• Moving average method
• Prediction performance
• Slope engineering
• Three gorges reservoir area

Affected by the rainfall and the periodic scheduling of reservoir water, some of the landslides in Three Gorges Reservoir area show the obvious step-like deformation. According to the deformation characteristics of the step-like landslides, a GA-SVM model for displacement prediction based on the response analysis of inducing factors was proposed. The landslide displacement was decomposed into trend item and periodic item by the moving average method, and the trend term displacement was predicted by the polynomial function, which is controlled by geological conditions of the landslide. As the periodic displacement is controlled by the combined action of all the inducing factors, the dominant factors were selected to establish the GA-SVM model for landslide displacement prediction. The total prediction displacement was the addition of the two items. The typical step-like landslide-Bazimen landslide-was taken as an example to make a displacement prediction by the GA-SVM model. The results showed that, the response analysis of inducing factors was the essentiality of landslide displacement prediction. The GA-SVM model based on the response of inducing factors showed great prediction performance in the step-like deformation, which is an effective displacement prediction method for the step-like landslide.

• 出版日期2015-9-15