支持向量机在边坡稳定分析预测的应用  

Application of a Support Vector Machine for Analysis and Prediction of Slope Stability

作　　者：洪勇[1,2] 邵珠山 马力[3] 

机构地区：[1]西安建筑科技大学材料与矿资学院,陕西西安710055 [2]西安建筑科技大学土木工程学院,陕西西安710055 [3]西安科技大学能源学院,陕西西安710054

出　　处：《沈阳建筑大学学报（自然科学版）》2017年第6期1004-1010,共7页Journal of Shenyang Jianzhu University：Natural Science

基　　金：国家自然科学基金项目(10772143);陕西省科技统筹项目(2015TZC-G-8-9)

摘　　要：目的提出一种基于改进算法的支持向量机模型(PSO-SVM),利用边坡的参数分析预测边坡稳定性.方法利用支持向量机有效解决小样本、高维数、非线性等问题的优势,建立粒子群算法(PSO)优化的支持向量机模型,粒子群算法优化支持向量机参数,模型中边坡几何参数和强度参数:边坡角β、边坡高度H、岩石容重γ、黏聚力c、内摩擦角φ以及孔隙水压力ru作为输入参数,边坡稳定性系数FS和边坡稳定状态S作为输出参数.结果 PSO-SVM模型与网格搜索算法(GS)、遗传算法(GA)优化SVM模型以及人工神经网络ANN模型相比,具有更高的分类精度和更强的预测能力.结论 PSO-SVM模型能够准确地获得边坡的稳定性系数,评价其稳定性,在边坡稳定分析和预测中具有良好的实际应用价值.A slope stability analysis and prediction model based on PSO-SVMalgorithm,which uses parameters of slopes,is put forward. As support vector machine（ SVM） could effectively solve the small sample,high dimension and non-linear problems,the Particle Swarm Optimization and Support Vector Machine（ PSO-SVM） coupling model was proposed to analyze the slope stability.The parameters of SVMwas optimized by the PSO. The most influential factors on the stability of slope such as slope angle,slope height,unit weight,cohesion,friction angle and pore water pressure coefficient were considered as model inputs,the factor of safety（ FS） or the stability status（ S） were considered as model outputs. Compared the results of case with optimizing SVMbased on grid search（ GS）,genetic algorithm（ GA） and artificial neural networks（ ANN） in details,theresults shows that PSO-SVMhas the higher classification accuracy and greater prediction ability among the four algorithms,and is therefore considered most suitable for slope stability analysis. it can acquire slope safety factors and evaluate slope stability,which can be well applied to the analysis of slope stability.

关 键 词：边坡稳定 粒子群算法 支持向量机 预测 

分 类 号：TU457[建筑科学—岩土工程]