Scholarlink

<p>Inclined drops (ID) are often used as energy dissipators in hydraulic structures. This study aims to evaluate the effectiveness of different machine learning techniques such as Gene Expression Programming (GEP), Multi-Layer Perceptron (MLP), Random Forest (RF), Radial Basis Functions (RBF) and Support Vector Machines (SVM) in predicting downstream relative depth (y<sub>d</sub>/h), relative energy loss (∆E/E<sub>u</sub>) and Froude number downstream (Fr<sub>d</sub>) in Inclined Gabion Drops (IGD). In each model, the relative critical depth and the inclination angle within the ID are used as parameter inputs. The results show that all models effectively estimate energy loss in ID structures. MLP outperformed the other models in terms of accuracy. MLP exhibited exceptional performance metrics with an R<sup>2</sup> value of 0.971, a DC coefficient of 0.971, an RMSE of 0.015, and a MAPE of 2.04. These metrics demonstrate the superior accuracy of MLP in predicting energy loss in IGD structures. Sensitivity analysis revealed that the y<sub>c</sub>/h parameter has a greater impact on the modeling process compared to other hydraulic parameters and energy loss in the IGD structure. This indicates that the relationship between y<sub>c</sub> (critical depth) and h (height of the drop) strongly influences the overall performance of the model.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Exploring machine learning models for assessing energy dissipation on inclined gabion drops

  • 期刊论文

  • 发表日期：

    2025年9月12日

查看全文
  • Mohammad Bagherzadeh,
  • Mirali Mohammadi,
  • Rasoul Daneshfaraz,
  • Amir Ghaderi,
  • Mohammad Reza Nikoo,
  • Alban Kuriqi

摘要

Inclined drops (ID) are often used as energy dissipators in hydraulic structures. This study aims to evaluate the effectiveness of different machine learning techniques such as Gene Expression Programming (GEP), Multi-Layer Perceptron (MLP), Random Forest (RF), Radial Basis Functions (RBF) and Support Vector Machines (SVM) in predicting downstream relative depth (yd/h), relative energy loss (∆E/Eu) and Froude number downstream (Frd) in Inclined Gabion Drops (IGD). In each model, the relative critical depth and the inclination angle within the ID are used as parameter inputs. The results show that all models effectively estimate energy loss in ID structures. MLP outperformed the other models in terms of accuracy. MLP exhibited exceptional performance metrics with an R2 value of 0.971, a DC coefficient of 0.971, an RMSE of 0.015, and a MAPE of 2.04. These metrics demonstrate the superior accuracy of MLP in predicting energy loss in IGD structures. Sensitivity analysis revealed that the yc/h parameter has a greater impact on the modeling process compared to other hydraulic parameters and energy loss in the IGD structure. This indicates that the relationship between yc (critical depth) and h (height of the drop) strongly influences the overall performance of the model.