Application of a regression model for predicting traffic volume from dynamic monitoring data to the bridge safety evaluation
Application of a regression model for predicting traffic volume from dynamic monitoring data to the bridge safety evaluation
复制标题
DOI:
10.1007/s13349-017-0234-7
复制
发表时间:
2017-08
影响因子:
4.4
通讯作者:
Kaiwan Wattana;M. Nishio
中科院分区:
文献类型:
--
作者:
Kaiwan Wattana;M. Nishio
This paper presents the verification and application of estimated traffic volume obtained from a statistical model. The model was constructed from dynamic responses acquired by a structural health monitoring (SHM) system installed on an in-service cable-stayed bridge. The SHM system consists of accelerometers, temperature sensors, and a traffic-counting system based on installed cameras on the bridge. The model performance was firstly assessed, and it was concluded that the constructed regression model for estimating the number of equivalent trucks from the dynamic responses was applicable. However, it was also recognized that the traffic conditions such as the number of passing vehicles and the speed of traffic flow slightly affected the estimation accuracies. Then, the relationships between the traffic volume and the adopted dynamic responses in the statistical model were verified by using the finite element (FE) model in order to apply in the bridge safety evaluation. The analytical results from the FE model revealed that the adopted responses obtained from the analytical data affected by only the traffic volume variability corresponded with those from the measurement, i.e., the adopted responses were sensitive to the traffic volume. Therefore, they were valid for use in the estimation of traffic volume. After that, the reliability indices of the target bridge were calculated by using the estimated traffic volume as the operational load effects with comparison of those calculated from the designed load in the design standard. The bridge reliability based on the estimated traffic volume could provide the actual safety conditions under the operational load, and it contributed more to decision making in bridge maintenance than those based on the design load.