基于支持向量机的边坡稳定性预测模型  

SUPPORT VECTOR MACHINE MODEL IN SLOPE STABILITY EVALUATION

作　　者：罗战友[1] 杨晓军[1] 龚晓南[1] 

机构地区：[1]浙江大学岩土工程研究所

出　　处：《岩石力学与工程学报》2005年第1期144-148,共5页Chinese Journal of Rock Mechanics and Engineering

摘　　要：根据影响边坡稳定性的主要因素,建立了边坡稳定性的支持向量机预测模型。该模型通过有限的经验数据的学习,建立了边坡稳定性与其影响因素之间的非线性关系。运用所建立的模型对具体的岩体边坡进行了判定,由结果知,基于线性核的支持向量机分类器不能有效地建立边坡稳定与影响因素之间的非线性映射,而基于神经网络核及径向基函数核的分类器能正确判定边坡的稳定性。Based on the main factors with important influence on slope stability, the support vector machine (SVM) model of slope evaluation is established. The nonlinear relation between slope stability and influencing factors is obtained from the finite empirical data by SVM model, and the model is applied to the practical engineering. Based on the results, it is not effective enough for SVM model of linear kernel function to find the nonlinear mapping between classification of slope stability and influencing factors, but the classifying device based on neural kernel functions and radial based function (RBF) kernel can correctly determine the classification of slope stability.

关 键 词：岩土力学 边坡稳定 统计学习 支持向量机 核函数 

分 类 号：TU457[建筑科学—岩土工程]