Scholarlink

Background <p>The impacts of wind-induced interference arise due to nearby structures, leading to changes in wind loads on a building.</p> Purpose <p>These changes depend primarily on the configuration and placement of surrounding buildings, their alignment concerning the wind direction, and their exposure in the upstream direction.</p> Methods <p>In this research, numerical simulations were performed to analyze the impact of wind loads on low-rise buildings and investigate the mean wind-induced interference effects between these structures. The goal is to propose practical methods and measures for wind-resistant design based on these findings. In recent times, machine learning (ML) has proven effective in predicting wind pressure coefficients. Despite their accuracy, ML models often lack the ability to instill confidence in end-users due to the inherent nature of predictions. This study utilizes the machine learning tool especially, the XG Boost tree-based regression model to forecast the interference factor (IF) and interference difference (ID) for low-rise mono-slope roof buildings.</p> Results <p> The study results indicate that greater interference effects were observed on the roofs of low-rise buildings situated at the corners or outer edges of the groups. Additionally, the XG Boost algorithm accurately predicted values for the interference factor (IF) and interference difference (ID),&#xa0;closely aligning with the values derived from computational fluid dynamics (CFD) simulations.</p> Conclusions <p>The XG Boost algorithm's capacity to generalize even when confronted with limited&#xa0;data makes it an appealing tool for acquiring knowledge on wind-induced interference effects,&#xa0;especially in situations where there is currently no existing theory or empirical generalization.</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

Extrapolating Wind-Induced Interference on Mono-Slope Roof of Low-Rise Buildings: CFD Simulation vs XG Boost Algorithm

  • 期刊论文

  • 发表日期：

    2026年1月7日

查看全文
  • Deepak Sharma,
  • Shilpa Pal,
  • Ritu Raj,
  • Nipun Bansal

摘要

Background

The impacts of wind-induced interference arise due to nearby structures, leading to changes in wind loads on a building.

Purpose

These changes depend primarily on the configuration and placement of surrounding buildings, their alignment concerning the wind direction, and their exposure in the upstream direction.

Methods

In this research, numerical simulations were performed to analyze the impact of wind loads on low-rise buildings and investigate the mean wind-induced interference effects between these structures. The goal is to propose practical methods and measures for wind-resistant design based on these findings. In recent times, machine learning (ML) has proven effective in predicting wind pressure coefficients. Despite their accuracy, ML models often lack the ability to instill confidence in end-users due to the inherent nature of predictions. This study utilizes the machine learning tool especially, the XG Boost tree-based regression model to forecast the interference factor (IF) and interference difference (ID) for low-rise mono-slope roof buildings.

Results

The study results indicate that greater interference effects were observed on the roofs of low-rise buildings situated at the corners or outer edges of the groups. Additionally, the XG Boost algorithm accurately predicted values for the interference factor (IF) and interference difference (ID), closely aligning with the values derived from computational fluid dynamics (CFD) simulations.

Conclusions

The XG Boost algorithm's capacity to generalize even when confronted with limited data makes it an appealing tool for acquiring knowledge on wind-induced interference effects, especially in situations where there is currently no existing theory or empirical generalization.