基于XGBoost的初始地应力场修正反演法
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TV22

基金项目:

国家自然科学基金项目(51779084)


A modified inversion method of initial geo-stress field based on XGBoost algorithm
Author:
Affiliation:

Fund Project:

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

    为解决多元线性回归反演地应力场可能存在的非线性拟合不足问题,引入极端梯度提升(XGBoost)算法构建数学模型,对多元线性回归的反演结果进行修正,获得更加准确的初始地应力场。实际工程应用表明:通过地应力测点的修正反演值和实测值的对比分析,测试集的拟合误差降低约60%,且稳定性高,有效提高了反演结果的精度。

    Abstract:

    To solve the problem of insufficient non-linear fitting that may exist in the inversion of geo-stress field using the multiple linear regression method, a mathematical model based on the extreme gradient boosting (XGBoost) algorithm is introduced, which can correct the inversion results of the multiple linear regression, and a more accurate initial geo-stress field can be obtained. Application to actual engineering cases show that, through the comparison and analysis of the corrected inversion value and the measured value of the ground stress measuring points, the fitting error of the test set is reduced by about 60%, and the stability is high, which effectively improves the accuracy of the inversion results.

    参考文献
    相似文献
    引证文献
引用本文

刘敬文,白金朋,王建,等.基于XGBoost的初始地应力场修正反演法[J].水利水电科技进展,2022,42(2):101-108.(LIU Jingwen,BAI Jinpeng,WANG Jian,et al. A modified inversion method of initial geo-stress field based on XGBoost algorithm[J]. Advances in Science and Technology of Water Resources,2022,42(2):101-108.(in Chinese))

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:2021-01-25
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2022-03-09
  • 出版日期: