Scholarlink

This study offers a novel approach to examining the dynamic interplay between a moving pedestrian and a flexible structure (a lab-scaled footbridge). Plantar capacitive sensors (Novel GmbH Loadsol®) are used to detect dynamic magnitudes (Ground Reaction Forces, or GRF), and inertial measurement units (Movella® Xsens DOT) are used to measure kinematic magnitudes (acceleration). The video from the standard camera and the data from the wireless sensors (which were first created for biomechanical applications) must be synchronized. To determine the precise timing of each footstep event and its spatial location, the data collected are combined with video frames of the pedestrian walking, and deep learning and image recognition techniques are used. This method makes it possible to create a realistic pedestrian loading model that can be used to assess the interaction effects between the pedestrian and the flexible structure. The paper also discusses possible applications (such as dynamic identification or Finite Element Model Updating)

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

Using Wearable Sensors and Image Recognition Techniques to Develop a Realistic Walking Load Model

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • David Fuentes-Jiménez,
  • Guillermo Fernández,
  • Alvaro Iglesias-Pordomingo,
  • Emiliano Pereira-González,
  • Pedro Gil-Jiménez,
  • Antolin Lorenzana

摘要

This study offers a novel approach to examining the dynamic interplay between a moving pedestrian and a flexible structure (a lab-scaled footbridge). Plantar capacitive sensors (Novel GmbH Loadsol®) are used to detect dynamic magnitudes (Ground Reaction Forces, or GRF), and inertial measurement units (Movella® Xsens DOT) are used to measure kinematic magnitudes (acceleration). The video from the standard camera and the data from the wireless sensors (which were first created for biomechanical applications) must be synchronized. To determine the precise timing of each footstep event and its spatial location, the data collected are combined with video frames of the pedestrian walking, and deep learning and image recognition techniques are used. This method makes it possible to create a realistic pedestrian loading model that can be used to assess the interaction effects between the pedestrian and the flexible structure. The paper also discusses possible applications (such as dynamic identification or Finite Element Model Updating)