Scholarlink

<p>In the context of global warming, the frequency of “grey swan” typhoons affecting China's southeastern coastal areas is increasing, posing a significant threat of collapse to tower cranes at coastal construction sites. To reduce the risk of accidents, it is crucial to implement real-time monitoring and response prediction for tower crane structures. This allows for the timely tracking of structural displacement trends and provides sufficient time for site emergency response. This paper aims to propose a hybrid data-model approach for predicting extreme typhoon-induced tower crane responses. Using a tower crane in a real project as an example, this paper developed an <i>IoT</i>-based real-time displacement monitoring system to capture the crane's load response data throughout the construction period. Additionally, a finite element model was developed to determine the crane's tower displacement under various extreme wind loads, compensating for the lack of extreme samples in the measured data. Subsequently, the CNN-BiLSTM-AdaBoost algorithm was used to predict the structural displacements of the tower crane under constant wind effects and extreme typhoon effects, employing both the data-driven method and the hybrid data-model approach. The results indicate that the pure data-driven method has limitations and may significantly underestimate the tower's structural displacement under extreme typhoon effects. In contrast, the hybrid data-model method effectively integrates real-time monitoring data with extreme data from finite element analysis, enhancing the accuracy and reliability of tower displacement predictions under extreme typhoon effects.</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

A hybrid data-model driven approach for predicting extreme typhoon-induced tower crane responses

  • 期刊论文

  • 发表日期：

    2025年5月29日

查看全文
  • Qiang Li,
  • Peixuan Wang,
  • Jun Zhang,
  • Hairong Wu,
  • Ping Wu,
  • Mingfeng Huang,
  • Dongming Lu

摘要

In the context of global warming, the frequency of “grey swan” typhoons affecting China's southeastern coastal areas is increasing, posing a significant threat of collapse to tower cranes at coastal construction sites. To reduce the risk of accidents, it is crucial to implement real-time monitoring and response prediction for tower crane structures. This allows for the timely tracking of structural displacement trends and provides sufficient time for site emergency response. This paper aims to propose a hybrid data-model approach for predicting extreme typhoon-induced tower crane responses. Using a tower crane in a real project as an example, this paper developed an IoT-based real-time displacement monitoring system to capture the crane's load response data throughout the construction period. Additionally, a finite element model was developed to determine the crane's tower displacement under various extreme wind loads, compensating for the lack of extreme samples in the measured data. Subsequently, the CNN-BiLSTM-AdaBoost algorithm was used to predict the structural displacements of the tower crane under constant wind effects and extreme typhoon effects, employing both the data-driven method and the hybrid data-model approach. The results indicate that the pure data-driven method has limitations and may significantly underestimate the tower's structural displacement under extreme typhoon effects. In contrast, the hybrid data-model method effectively integrates real-time monitoring data with extreme data from finite element analysis, enhancing the accuracy and reliability of tower displacement predictions under extreme typhoon effects.