摘要
边坡稳定性分析是岩土工程的一个常见问题,影响参数较多。首先将影响边坡稳定性的样本集合建立支持向量机(Support Vector Machine,SVM)回归模型,而后使用网格搜索法(Grid-search)优化支持向量机的参数,并将优化过参数的支持向量机回归模型与贝叶斯岭回归模型、普通线性回归模型、梯度增强回归模型的预测结果进行对比。研究结果表明:优化后的SVM回归模型预测方法在边坡安全系数预测方法中更为精准稳定,具有一定的实际应用价值。
Slope stability analysis is a common problem with many inf luencing parameters in geotechnical engineering.At First,the sample set that affects the slope stability is built into the support vector machine (SVM) regression model.Then,the grid-search method is used to optimize the parameters of the support vector machine.Next,the support vector machine regression model with optimized parameters is compared with Bayesian Ridge Regression model,the general linear regression model and the gradient-enhanced regression model in terms of prediction results.The results show that the optimized SVM regression model prediction method is more accurate and stable in comparison with other slope safety factor prediction methods.This method is more practicable.
作者
王健伟
徐玉胜
李俊鑫
WANG Jianwei;XU Yushen;LI Junxing(Postgraduate Department,China Academy of Railway Sciences,Beijing 100081;China Academy of Railway Sciences (Shenzhen)Research and Design Institute Co. Ltd.,Shenzhen Guandong 518034,China;Shenzhen Geological Disaster MonitoringEngineering Laboratory,Shenzhen Guangdong 518000,China)
出处
《铁道建筑》
北大核心
2019年第5期94-97,共4页
Railway Engineering
基金
中国铁道科学研究院基金(2016YJ156)
关键词
边坡安全系数
支持向量机
统计分析
边坡
机器学习
Slope safety factor
Support vector machine
Statistical analysis
Slope
Machine learning
