摘要
为了实现斜拉桥缆索的自动无损检测,针对斜拉桥缆索自动无损检测方法进行研究,提出基于蛇形机器人多传感器数据融合的缆索缺陷自动检测方法。通过搭载多传感器的蛇形机器人螺旋攀爬运动,实现在役自动检测;利用数据融合技术对多传感器的数据进行融合实现桥梁缆索缺陷的自动检测,在数据层采用加权平均进行信号融合,在特征层采用支持向量机作为缺陷分类与识别平台,在决策层应用D-S证据理论对缺陷做出最后决策,并应用有限元软件ANASYS建立缆索缺陷仿真模型,在MATLAB上对数据融合算法进行验证。研究结果表明该方法能够方便地实现缆索缺陷在役自动检测,不仅可以降低系统的不确定性,而且能有效地提高缆索缺陷识别精度和可靠性。
In order to achieve automatic non the automatic non-destructive testing method -destructive testing of cable-stayed bridge cables, we researched of cable-stayed bridge, and put forward the automatic detecting method of cable defects based on snake-like robot and multi-sensor data fusion. Through the spiral climbing motion of snake-like robot with multi sensors, we realized inservice automation detection. Based on data fusion technology, using weighted average signal fusion in the data layer, support vector machine as defect classification and identification platform in the characteristic layer, and D-S evidence theory to make final decision making in the decision layer, we realized automatic detection of bridge cable defects by multi-sensor data fusion. Then, we used ANASYS to establish cable defect simulation model, and used MATLAB to verify the data fusion algorithm. The result shows that the presented method could easily implement automatic detection of inservice cable defects, it not only could greatly reduce the uncertainty of the system, but also could effectively improve the defect identification accuracy and reliability.
出处
《公路交通科技》
CAS
CSCD
北大核心
2011年第12期88-93,共6页
Journal of Highway and Transportation Research and Development
基金
交通运输部西部交通建设科技项目(B1110210)
关键词
桥梁工程
自动无损检测
数据融合
缆索缺陷检测
蛇形机器人
bridge engineering
automatic non-destructive testing
data fusion
cable defect detection
snake-like robot
