论文部分内容阅读
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 realize the automatic non-destructive testing of cables of cable-stayed bridge, an automatic non-destructive testing method for cables of cable-stayed bridges is studied. An automatic detection method of cable defects based on multisensor data fusion of snake robots is proposed. The multi-sensor snake-like robots are used to carry out the spiral climbing motion to realize the automatic detection in service. The data fusion technology is used to fuse the multi-sensor data to automatically detect the bridge cable defects. The weighted average is used for signal fusion in the data layer. At the decision-making level, DS evidence theory was used to make the final decision on defects. The finite element software ANASYS was used to establish the cable defect simulation model, and the data fusion algorithm was verified on MATLAB. The results show that this method can be used to automatically detect the defects of cables in service, which can not only reduce the system uncertainty but also effectively improve the accuracy and reliability of cable defect identification.