新型路基沉降预测方法及其展望  

New Roadbed Settlement Prediction Methods and Their Outlook

作　　者：廖向阳 LIAO Xiang-yang(Hunan Transportation planning Survey and Design Institute Co.,Ltd.,Changsha 410000,China)

机构地区：[1]湖南省交通规划勘察设计院有限公司,长沙410000

出　　处：《价值工程》2023年第19期166-168,共3页Value Engineering

摘　　要：传统路基沉降预测方法非线性映射能力差。提出一种基于机器学习的路基沉降预测模型,并以江门市新会区银鹭大道路基工程为例,验证该预测模型的有效性。结果表明:所提预测模型被验证效果较好,各模型预测性能排序为:SVR模型>BP模型>RF模型;图形神经网络在路基沉降预测领域的应用具有良好发展前景。研究成果可为路基沉降的预测提供理论参考。The traditional roadbed settlement prediction method has poor nonlinear mapping capability.A machine learning based roadbed settlement prediction model is proposed and the effectiveness of the prediction model is verified by taking the roadbed project of Yinlu Avenue in Xinhui District,Jiangmen City as an example.The results showed that the proposed prediction model was validated to be effective,and the prediction performance of each model was ranked as:SVR model>BP model>RF model;the application of graphical neural network in the field of roadbed settlement prediction has good development prospects.The research results can provide theoretical reference for the prediction of roadbed settlement.

关 键 词：交通线路 路基沉降 沉降预测方法 机器学习 

分 类 号：U416.1[交通运输工程—道路与铁道工程]