基于支持向量机的多年冻土路基融沉变形预测

发布时间：2018-02-04 08:29

本文关键词： 多年冻土路基融沉变形支持向量机预测模型对比分析　出处：《公路工程》2014年05期 　论文类型：期刊论文

【摘要】：多年冻土路基热稳定性差,工后沉降量大,路基病害较为严重。如果能够准确预测该类路基的工后融沉值,就可为工程建设提供重要的参考依据,从而提高该类路基的路用性能。为此,在对现有预测模型应用效果分析的基础上,首次将支持向量机应用于多年冻土路基融沉变形的预测中,提出了一种有效可行的新型预测方法,并以实际工程为依托,构建了基于支持向量机原理的多年冻土路基融沉变形预测模型。通过与实测值及其它预测模型的对比分析表明:该模型在预测过程中有效的避免了"过拟合"及"维数灾难",人为干预较少,具有预测精度高,泛化能力强,预测结果稳定的特点,成功的解决了多年冻土路基影响因素多,样本数量少等带来的预测难题。
[Abstract]:The thermal stability of permafrost subgrade is poor, the amount of settlement after construction is large, and the disease of roadbed is more serious. If we can accurately predict the value of post-construction thaw settlement of this kind of subgrade, it can provide an important reference basis for engineering construction. In order to improve the roadway performance of this kind of subgrade, based on the analysis of the application effect of the existing prediction model, support vector machine (SVM) is applied to the prediction of thawing settlement of permafrost subgrade for the first time. A new effective and feasible prediction method is proposed, which is based on the actual project. A prediction model of thawing settlement of permafrost subgrade based on support vector machine (SVM) is constructed. The comparison with the measured values and other prediction models shows that the model can effectively avoid "overfitting" in the prediction process. And "Dimension disaster". The artificial intervention is less, the prediction accuracy is high, the generalization ability is strong, and the forecast result is stable. It has successfully solved the difficult problem of prediction brought by the influence factors of permafrost roadbed and the small number of samples.
【作者单位】：东北林业大学土木工程学院;
【分类号】：U416.168
【正文快照】： 我国是世界第三大冻土国,多年冻土分布面积约为2.15×106km2,占国土面积的22.3%,主要分布在东北的大小兴安岭地区、西部的西藏高山及青藏高原等地[1],由于多年冻土区多为经济欠发达地区,为改善当地的经济环境,公路建设首当其冲。然而,在多年冻土区修建公路,由于冻土的不断退融

【参考文献】