| 注册
首页|期刊导航|人民珠江|基于WOA-RFR的混凝土坝变形预测监控模型

基于WOA-RFR的混凝土坝变形预测监控模型

冯瑜 吴云星 谷汶静 庞琼 谷艳昌 陈斯煜

人民珠江2024,Vol.45Issue(7):118-124,7.
人民珠江2024,Vol.45Issue(7):118-124,7.DOI:10.3969/j.issn.1001-9235.2024.07.014

基于WOA-RFR的混凝土坝变形预测监控模型

Prediction and Monitoring Model of Concrete Dam Deformation Based on WOA-RFR

冯瑜 1吴云星 2谷汶静 3庞琼 2谷艳昌 2陈斯煜2

作者信息

  • 1. 黄山市水利水电建设管理站,安徽 黄山 245000
  • 2. 南京水利科学研究院大坝安全与管理研究所,江苏 南京 210029||水利部大坝安全管理中心,江苏 南京 210029
  • 3. 华北水利水电大学水资源学院,河南 郑州 450046
  • 折叠

摘要

Abstract

The random forest algorithm and whale optimization algorithm were introduced in the construction of the prediction model of concrete dam deformation based on WOA-RFR to improve the prediction accuracy and model performance.The random forest model belonging to the machine learning algorithm has many advantages such as strong generalization ability and fast training speed,and it has a strong mapping capability for nonlinear features.However,because different parameters and corresponding parameter combinations of the primitive random forest algorithm have a great influence on the improvement and stability of the model performance,the effectiveness of the results cannot be guaranteed under the manual empirical method.Therefore,to address the parameter calibration of the random forest model,the whale optimization algorithm with strong global search ability is introduced to conduct combination optimization on key parameters.The aim is to further enhance the model's generalization ability and robustness at the same time as obtaining optimal parameter combinations.The monitoring model of dam deformation is built by using the random forest optimized by whale algorithm for an actual project,and the coefficient of determination,root mean square error(RMSE),and mean absolute percentage error(MAPE)are introduced to evaluate and compare the excellent performance of the proposed models.The prediction results were compared with different intelligent optimization algorithms and multiple control models.The results show that the WOA-RFR prediction model has higher prediction accuracy and stability,and WOA optimization significantly improves the model performance.

关键词

混凝土坝/大坝变形预测/随机森林模型/鲸鱼优化算法

Key words

concrete dam/dam deformation prediction/random forest model/whale optimization algorithm

分类

建筑与水利

引用本文复制引用

冯瑜,吴云星,谷汶静,庞琼,谷艳昌,陈斯煜..基于WOA-RFR的混凝土坝变形预测监控模型[J].人民珠江,2024,45(7):118-124,7.

基金项目

南京水科院基本科研业务费科研创新团队建设项目(Y722003) (Y722003)

国家自然科学基金项目(52309157) (52309157)

南京水科院基本科研业务费重点项目(Y723002) (Y723002)

人民珠江

1001-9235

访问量9
|
下载量0
段落导航相关论文