Dynamic load data have been Valuable for bridge and pavement design. Traditional ways to acquire truck axle and gross weight information are expensive and subject to bias, and this has led to the development of Weigh-in-Motion (WIM) techniques. Most of the existing WIM systems have been developed to measure only the static axle loads. This paper aims to introduce a method to identify moving dynamic loads on bridges using the bridge responses caused by such loads. A closed-form solution can be obtained to identify moving constant loads, while numerical methods must be used to identify the time-varying moving loads. The set of equations that leads to the solution is based on Euler's equation of beams. A two-axle vehicle model is developed to generate the theoretical responses and the corresponding interactive moving forces. Studies were carried out to check whether the proposed method could recover the original interactive moving forces and the method was found to be feasible.

An interpretive method for moving force identification

For structural health monitoring (SHM) of civil infrastructures, displacement is a good descriptor of the structural behavior under all the potential disturbances. However, it is not easy to measure displacement of civil infrastructures, since the conventional sensors need a reference point, and inaccessibility to the reference point is sometimes caused by the geographic conditions, such as a highway or river under a bridge, which makes installation of measuring devices time-consuming and costly, if not impossible. To resolve this issue, a visionbased real-time displacement measurement system using digital image processing techniques is developed. The effectiveness of the proposed system was verified by comparing the load carrying capacities of a steel-plate girder bridge obtained from the conventional sensor and the present system. Further, to simultaneously measure multiple points, a synchronized vision-based system is developed using master/slave system with wireless data communication. For the purpose of verification, the measured displacement by a synchronized vision-based system was compared with the data measured by conventional contact-type sensors, linear variable differential transformers (LVDT) from a laboratory test.

Development and Application of a Vision-based Displacement Measurement System for Structural Health Monitoring of Civil Structures

A multi-resolution deep feature framework for dynamic displacement measurement of bridges using vision-based tracking system

We investigate the capabilities of transfer learning in the area of structural health monitoring. In particular, we are interested in damage detection for concrete structures. Typical image datasets for such problems are relatively small, calling for the transfer of learned representation from a related large-scale dataset. Past efforts of damage detection using images have mainly considered cross-domain transfer learning approaches using pre-trained ImageNet models that are subsequently fine-tuned for the target task. However, there are rising concerns about the generalizability of ImageNet representations for specific target domains, such as for visual inspection and medical imaging. We, therefore, evaluate a combination of in-domain and cross-domain transfer learning strategies for damage detection in bridges. We perform comprehensive comparisons to study the impact of cross-domain and in-domain transfer, with various initialization strategies, using six publicly available visual inspection datasets. The pre-trained models are also evaluated for their ability to cope with the extremely low-data regime. We show that the combination of cross-domain and in-domain transfer persistently shows superior performance specially with tiny datasets. Likewise, we also provide visual explanations of predictive models to enable algorithmic transparency and provide insights to experts about the intrinsic decision logic of typically black-box deep models.

/pdf/damage-detection-using-in-domain-and-cross-domain-transfer-5sxb6e94za.pdf

Damage detection using in-domain and cross-domain transfer learning

A review of the research and application of deep learning-based computer vision in structural damage detection

Manual visual inspection is customarily used to identify and evaluate service status of bridges. However, current procedures followed by human inspectors demand long inspection time to access bridg...

Vision-based defects detection for bridges using transfer learning and convolutional neural networks

Information on dynamic displacement is an effective indicator for structures condition evaluation and provides a quantised insight into the structural analysis. This paper proposes a low-cost syste...

A marker-free method for structural dynamic displacement measurement based on optical flow

An accurate approach for obtaining spatiotemporal information of vehicle loads on bridges based on 3D bounding box reconstruction with computer vision

This paper investigates the assessment of corrosion effects on the performance of the composite bridge caused by chloride ions during the service life depending on Cellular Automata (CA) and Finite...

Jinsong Zhu

Papers

Vision-based defects detection for bridges using transfer learning and convolutional neural networks

A multi-resolution deep feature framework for dynamic displacement measurement of bridges using vision-based tracking system

A marker-free method for structural dynamic displacement measurement based on optical flow

An accurate approach for obtaining spatiotemporal information of vehicle loads on bridges based on 3D bounding box reconstruction with computer vision

Evaluation of the corrosion effects on the performance of composite bridge based on cellular automata and finite element method