Collecting the information of traffic load, especially heavy trucks, is crucial for bridge statistical analysis, safety evaluation, and maintenance strategies. This paper presents a traffic sensing methodology that combines a deep learning based computer vision technique with the influence line theory. Theoretical background and derivations are introduced from both aspects of structural analysis and computer vision techniques. In addition, to evaluate the effectiveness and accuracy of the proposed traffic sensing method through field tests, a systematic analysis is performed on a continuous box-girder bridge. The obtained results show that the proposed method can automatically identify the vehicle load and speed with promising efficiency and accuracy and most importantly cost-effectiveness. All these features make the proposed methodology a desirable bridge weigh-in-motion system, especially for bridges already equipped with structural health monitoring system.

A reliable and accurate monitoring of traffic load is of significance for the operational management and safety assessment of bridges. Traditional weight-in-motion techniques are capable of identifying moving vehicles with satisfactory accuracy and stability, whereas the cost and construction induced issues are inevitable. A recently proposed traffic sensing methodology, combining computer vision techniques and traditional strain based instrumentation, achieves obvious overall improvement for simple traffic scenarios with less passing vehicles, but are enfaced with obstacles in complicated traffic scenarios. Therefore, a traffic monitoring methodology is proposed in this paper with extra focus on complicated traffic scenarios. Rather than a single sensor, a network of strain sensors of a pre-installed bridge structural health monitoring system is used to collect redundant information and hence improve accuracy of identification results. Field tests were performed on a concrete box-girder bridge to investigate the reliability and accuracy of the method in practice. Key parameters such as vehicle weight, velocity, quantity, type and trajectory are effectively identified according to the test results, in spite of the presence of one-by-one and side-by-side vehicles. The proposed methodology is infrastructure safety oriented and preferable for traffic load monitoring of short and medium span bridges with respect to accuracy and cost-effectiveness.

Mode shapes have been playing a vital role in the research and application of bridge structural health monitoring. This paper presents a novel indirect method identifying bridge mode shapes using dynamic responses of a tractortrailer vehicle model, which consists of one tractor and three instrumented trailers. In an effort to eliminate the road roughness effect, accelerations of adjacent trailers are firstly subtracted. Wavelet analysis is then employed to identify bridge mode shapes from the subtracted accelerations in an iterative manner. Furthermore, wavelet denoising algorithm is adopted to improve the identification accuracy in the presence of measurement noise. Systematic numerical simulations, in which a tractor-trailer model passes over an expressway bridge, are conducted in order to investigate the performance of the proposed method. Sensitivity analysis including vehicle speed, class of road roughness, and noise level are studied in this numerical investigation. Results demonstrate that the proposed method is able to identify bridge modal frequencies and mode shapes with satisfactory resolution, accuracy, and robustness.

Summary Over the last several decades, a lot of bridges have been equipped with the bridge structural health monitoring system, leading to an accumulation of voluminous monitoring data. Since the sensors and associated transmission hardware are subjected to harsh environments, the monitoring data frequently contains various faults, and it is laborious to cleanse the data manually. For the purpose of automatically detecting and classifying faulty monitoring data in large quantities, this paper proposes a novel method that uses the relative frequency distribution histograms (RFDH) of monitoring data as well as the one‐dimensional convolutional neural network (1‐D CNN). The overall procedure of this method can be described as follows: First, RFDHs are constructed from different classes of hour‐long data segments. Second, inverted envelopes of the RFDHs are labeled as the training data to train the 1‐D CNN. Third, a well‐trained 1‐D CNN is used to detect and classify long‐term monitoring data according to their RFDHs of hour‐long data segments. Comprehensive validation of the proposed method is conducted with selective acceleration data collected from two long‐span bridges. The validation yields satisfactory results, demonstrating the accuracy, efficiency, and generality of the method.

Complicated traffic scenarios, including random change of vehicles’ speed and lane, as well as the simultaneous presence of multiple vehicles on bridge, are main obstacles that prevents bridge weigh-in-motion (BWIM) technique from reliable and accurate application. To tackle the complicated traffic problems of BWIM, this paper develops a novel BWIM method which integrates deep-learning-based computer vision technique and bridge influence surface theory. In this study, bridge strains and traffic videos are recorded synchronously as the data source of BWIM. The computer vision technique is employed to detect and track vehicles and corresponding axles from traffic videos so that spatio-temporal paths of vehicle loads on the bridge can be obtained. Then a novel method is proposed to identify the strain influence surface (SIS) of the bridge structure based on the time-synchronized strain signals and vehicle paths. After the SIS is identified, the axle weight (AW) and gross vehicle weight (GVW) can be identified by integrating the SIS, time-synchronized bridge strain, and vehicle paths. For illustration and verification, the proposed method is applied to identify AW and GVW in scale model experiments, in which the vehicle-bridge system is designed with high fidelity, and various complicated traffic scenarios are simulated. Results confirm that the proposed method contributes to improve the existing BWIM technique with respect to complicated traffic scenarios.

No abstract

No abstract

Accurate finite element (FE) models play an essential role in the health monitoring of operational bridges. Static structural deflections caused by vehicles, which are used in traditional finite element model updating (FEMU) methods, are often procured from field tests, interrupting the traffic and limiting the test loading scenarios. This study proposes a FEMU method that directly applies the massive, multi-source structural and traffic data in the operation phase to update the FE model, effectively solving the defects above. We use the computer vision-based vehicle load identification technique to accurately locate and weigh vehicle loads and carry out static simulations in the FE model based on the identified vehicle loads. The proposed FEMU objective function is established using indices including dynamic structural characteristics and vehicle-induced static structural responses. The static error index in the objective function integrates the curve shape and extrema difference of theoretic and measured static responses. Finally, we deploy a parallel particle swarm optimization (PSO) algorithm to find the global optimal updated FE model. A continuous scale bridge model is employed in the experimental studies with four typical scenarios. Compared to the results from the initial FE model, the updated FE model provides significantly better results both in dynamic and static aspects in all scenarios, and the static error indexes reduce by 75% on average. The proposed method offers a practical approach to deploying the monitoring data for real-time FEMU, which considers dynamic and static features and provides a basis for damage detection, performance assessment, and management.