Scholarlink

<p>This paper presents the development of a time-synchronised wireless vision sensor network using the global navigation satellite system (GNSS). The sensor network, named the flexible vision network (FVN), offers significant advantages over existing wired or wireless time-synchronised vision sensor networks. These advantages include: 1) spatial flexibility, with no distance limitations between sensor nodes imposed by Ethernet cables or WiFi communication, 2) scalability in the number of nodes due to its independent time-sync operation based on satellites without any time-sync interaction with other nodes, and 3) straightforward time synchronisation with other heterogeneous sensor systems, such as accelerometers or dynamic strain systems, due to its independent time-sync operation providing an absolute time reference. A series of four laboratory experiments and one field experiment was conducted to validate the FVN, followed by an application experiment for simultaneous input–output measurements. The first laboratory experiment measured the timestamping error between two identical FVN nodes triggered by a common signal, finding a standard deviation of 17 µs in the timestamping difference. The second laboratory experiment assessed the timestamping error between two identical nodes tracking the same moving target, revealing a maximum time difference of 3.05 ms with rolling shutter cameras and 0.34 ms with global shutter cameras. This indicates that camera hardware significantly contributes to the error. The third laboratory experiment demonstrated a maximum displacement measurement error at 1/37 pixels level. The fourth laboratory experiment involved measuring time-synchronised displacements of 25 points on a laboratory floor structure using six nodes. The fifth field experiment measured displacements at 12 points along a footbridge. In both the laboratory and field experiments, the identified modal parameters were consistent with those obtained from wired acceleration measurement systems. The final experiment demonstrated a successful application of the FVN for time-synchronised input–output measurements in live pedestrian positioning and structural displacement, enabling the estimation of influence lines. While the experimental results were promising, the FVN requires clear visibility of the sky, which is generally achievable in field experiments involving civil infrastructure.</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

GNSS time-synchronised wireless vision sensor network for structural health monitoring

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年5月8日

查看全文
  • Miaomin Wang,
  • Zuo Zhu,
  • Ki Young Koo,
  • James Brownjohn

摘要

This paper presents the development of a time-synchronised wireless vision sensor network using the global navigation satellite system (GNSS). The sensor network, named the flexible vision network (FVN), offers significant advantages over existing wired or wireless time-synchronised vision sensor networks. These advantages include: 1) spatial flexibility, with no distance limitations between sensor nodes imposed by Ethernet cables or WiFi communication, 2) scalability in the number of nodes due to its independent time-sync operation based on satellites without any time-sync interaction with other nodes, and 3) straightforward time synchronisation with other heterogeneous sensor systems, such as accelerometers or dynamic strain systems, due to its independent time-sync operation providing an absolute time reference. A series of four laboratory experiments and one field experiment was conducted to validate the FVN, followed by an application experiment for simultaneous input–output measurements. The first laboratory experiment measured the timestamping error between two identical FVN nodes triggered by a common signal, finding a standard deviation of 17 µs in the timestamping difference. The second laboratory experiment assessed the timestamping error between two identical nodes tracking the same moving target, revealing a maximum time difference of 3.05 ms with rolling shutter cameras and 0.34 ms with global shutter cameras. This indicates that camera hardware significantly contributes to the error. The third laboratory experiment demonstrated a maximum displacement measurement error at 1/37 pixels level. The fourth laboratory experiment involved measuring time-synchronised displacements of 25 points on a laboratory floor structure using six nodes. The fifth field experiment measured displacements at 12 points along a footbridge. In both the laboratory and field experiments, the identified modal parameters were consistent with those obtained from wired acceleration measurement systems. The final experiment demonstrated a successful application of the FVN for time-synchronised input–output measurements in live pedestrian positioning and structural displacement, enabling the estimation of influence lines. While the experimental results were promising, the FVN requires clear visibility of the sky, which is generally achievable in field experiments involving civil infrastructure.