基于随机森林的内河架空直立式码头损伤诱因反演模型研究*
DOI:
CSTR:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

国家自然科学基金项目(51508059);重庆市交通局科技项目(2020-08);重庆市基础研究与前沿探索项目(cstc2018jcyjAX0345);重庆市教委科学技术研究项目(KJQN201800739)


Back analysis model of damage incentives of inland river overhead vertical wharf based on random forest
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    为研究对码头结构损伤不良诱因作用的反演,在内河架空直立式码头物理模型上进行试验,分析码头在超限堆载和不规范靠泊两种损伤诱因作用下的应变规律。结合三峡库区架空直立式码头损伤诱因反演分析,提出基于随机森林的损伤诱因反演模型,并使用物理模型的试验数据对模型进行评估。结果表明:由物理模型试验得到靠江侧和靠岸侧基桩轴向应变分布规律,靠江侧轴向应变特征较靠岸侧明显,更适合作为反演输入;随机森林反演模型能够对物理模型损伤诱因的类型、位置、强度进行较为准确的反演,损伤类型的精度达到0.98,损伤位置的精度达到0.99,损伤强度的精度达到0.99。

    Abstract:

    To study the back analysis of the adverse inducement of wharf structure damage,we carry out a test on the physical model of the inland river overhead vertical wharf,and analyze the strain law of the wharf under two kinds of damage inducement,namely overloading and irregular berthing.Combining with the back analysis of the damage inducement of the overhead vertical wharf in the Three Gorges reservoir area,we put forward a back analysis model of the damage inducement based on random forest,and evaluate the model by the tests data of the physical model.The results show that the distribution law of axial strain of foundation pile near the river and near the shore are obtained by physical model test,and the axial strain characteristics near the river are more obvious than those near the shore,so it is more suitable for back analysis input source.The random forest back analysis model can accurately predict the type,position and intensity of the horizontal and vertical actions of the physical model.The accuracy of the damage type reaches 0.98,and the accuracy of the damage location reaches 0.99,moreover the accuracy of damage intensity reaches 0.99.

    参考文献
    相似文献
    引证文献
引用本文

周世良,熊 聪,柯春儒,等.基于随机森林的内河架空直立式码头损伤诱因反演模型研究*[J].水运工程,2023(7):53-59.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2023-07-13
  • 出版日期:
文章二维码