视频数据挖掘的非接触式桥梁振动监测

发布时间：2018-06-25 00:30

本文选题：桥梁健康监测 + 桥梁振动　；参考：《重庆大学学报》2017年09期

【摘要】：为克服传统桥梁监测技术施工难、成本高和观测点少等不足,利用视觉技术实现了非接触式桥梁振动状态监测。该方法包含视频数据采集、视频数据分析与结果可视化等3个环节。其中,视频数据分析环节又包含3个步骤:首先,对视频中的稳定像素点进行定位;然后,根据视频中稳定像素点的数值变化,由视频数据挖掘方法建立视频热噪声模型;最后,由视频热噪声模型计算出视频中桥梁各局部位置上的振动频率。实验结果表明,该方法可有效侦测到桥梁可见结构部件上出现的人眼不易察觉的微弱振动,并对桥梁不同局部位置上出现的振动频率进行同步定量测算,从而提供了一种便捷高效、成本合理、通用性好的桥梁结构健康监测新途径。
[Abstract]:In order to overcome the shortcomings of the traditional bridge monitoring technology, such as difficult construction, high cost and few observation points, the non contact bridge vibration state monitoring is realized by visual technology. This method includes 3 links: video data acquisition, video data analysis and result visualization. Among them, the video data analysis link contains 3 steps: first, video A video thermal noise model is established by the video data mining method. Finally, the vibration frequency of each part of the bridge in the video is calculated by the video thermal noise model. The experimental results show that the method can detect the visible structural parts of the bridge effectively. It appears that the weak vibration of the human eye is not easy to detect, and the vibration frequency of different parts of the bridge is measured synchronously and quantitatively, thus providing a new way to monitor the health of bridge structure with convenient and efficient, reasonable cost and good generality.
【作者单位】：浙江大学光电科学与工程学院现代光学仪器国家重点实验室;衢州学院机械工程学院;
【基金】：浙江省科技计划资助项目(2014C33011) 浙江省自然科学基金资助(LY15F020042) 国家自然科学基金资助(61502273) 校级科研启动经费资助项目(BSYJ201415) 校级科学研究基金项目(KY1303)~~
【分类号】：TP391.41;U446

【相似文献】