长江科学院院报 ›› 2021, Vol. 38 ›› Issue (7): 46-50.DOI: 10.11988/ckyyb.20200444

• 工程安全与灾害防治 • 上一篇    下一篇

基于贝叶斯优化LightGBM的大坝变形预测模型

高治鑫1, 包腾飞1,2,3, 李扬涛1, 王一兵1   

  1. 1.河海大学 水利水电学院,南京 210098;
    2.河海大学 水文水资源与水利工程科学国家重点实验室,南京 210098;
    3.三峡大学 水利与环境学院,湖北 宜昌 443002
  • 收稿日期:2020-05-16 修回日期:2020-09-26 出版日期:2021-07-01 发布日期:2021-07-08
  • 通讯作者: 包腾飞(1974-),男,湖北黄冈人,教授,博士,博士生导师,研究方向为水工结构及岩土工程安全监控。E-mail: baotf@hhu.edu.cn
  • 作者简介:高治鑫(1997-),女,山东泰安人,硕士研究生,研究方向为水工结构安全监控。E-mail:gaozhixin@hhu.edu.cn
  • 基金资助:
    国家重点研发计划项目(2018YFC1508603);国家自然科学基金项目 (51739003)

Dam Deformation Prediction Model Based on Bayesian Optimization and LightGBM

GAO Zhi-xin1, BAO Teng-fei1,2,3, LI Yang-tao1, WANG Yi-bing1   

  1. 1. College of Water Conservancy and Hydropower, Hohai University, Nanjing 210098, China;
    2. State Key Laboratory of Hydrology Water Resources and Hydraulic Engineering,Hohai University,Nanjing 210098,China;
    3. College of Hydraulic & Environmental Engineering, China Three Gorges University, Yichang 443002, China
  • Received:2020-05-16 Revised:2020-09-26 Online:2021-07-01 Published:2021-07-08

摘要: 为了解决大坝变形预测模型易陷入局部最优及不适用大规模数据等问题,采用一种快速高效的基于决策树的梯度提升框架LightGBM,并结合全局优化算法——贝叶斯优化进行大坝变形预测。为验证模型适用性,以两座实际混凝土坝工程为例分析,并与多元线性回归、支持向量回归机和多层神经网络等预测结果进行比较。结果表明,该模型均方根误差(RMSE)和平均绝对误差(MAE)等指标均优于其他方法,验证了该模型的可行性及优越性。LightGBM可对输入参数的重要性进行评估,对影响大坝变形的特征进行筛选,从而确定对大坝变形影响更显著的因素,为后续的安全评估工作提供参考。

关键词: 大坝变形预测, 贝叶斯优化, 梯度提升框架, 多元线性回归, 支持向量回归机, 多层神经网络

Abstract: Deformation as an intuitive monitoring indicator reflects the operation state of a dam. Existing intelligent methods for dam deformation prediction are prone to local optimum and are inapplicable in the case of large-scale data. In this paper, we combine a fast and efficient gradient boosting framework Light Gradient Boosting Machine (LightGBM) based on the decision tree with the global optimization algorithm, Bayesian optimization, to predict dam deformation. Taking two concrete dams as case study, we compared the modelling result with those of multiple linear regression, support vector regression, and multi-layer perceptron to verify the applicability of the present model. The RMSE and MAE of the proposed model are both superior to those of other methods, which manifests the feasibility and superiority of the model. In addition, LightGBM can evaluate the importance of input parameters and select the features that affect dam deformation, so as to determine the factors that have more significant influence on dam deformation and offer reference for following safety assessment.

Key words: dam deformation prediction, Bayesian optimization, LightGBM, multiple linear regression, support vector regression, multi-layer perceptron

中图分类号: