Due to unavoidable uncertainties, it is critical to perform probabilistic analysis when examining the safety of coupled train-bridge system based on the dynamic interaction analysis. This paper proposes a novel approach for probabilistic safety analysis of coupled train-bridge system using deep learning based surrogate model. Deep neural network embedded with convolutional neural network is designed and developed to construct the surrogate model substituting the 3D train-bridge interaction system model for reducing the computational efforts and efficiently predicting a series of dynamic indices of the coupled system. Due to the lack of explicit expression of the performance function, automatic differentiation is exploited to derive the surrogate model and further to realise the reliability and sensitivity analysis with the first-order method. The obtained reliability and sensitivity indices are compared to that resulting from the Monte Carlo simulation method with importance sampling and tail modelling. The proposed approach is applied on a high-speed train and bridge system, treating multiple main system parameters as random variables. Effects of the railway operational speed on the system reliability are investigated. The results show that the proposed approach can provide an efficient solution to the probabilistic safety analysis of coupled train-bridge system especially with small failure probability.

Seismic response prediction is a challenging problem and is significant in every stage during a structure’s life cycle. Deep neural network has proven to be an efficient tool in the response prediction of structures. However, a conventional neural network with deterministic parameters is unable to predict the random dynamic response of structures. In this paper, a deep Bayesian convolutional neural network is proposed to predict seismic response. The Bayes-backpropagation algorithm is applied to train the proposed Bayesian deep learning model. A numerical example of a three-dimensional building structure is utilized to validate the performance of the proposed model. The result shows that both acceleration and displacement responses can be predicted with a high level of accuracy by using the proposed method. The main statistical indices of prediction results agree closely with the results from finite element analysis. Furthermore, the influence of random parameters and the robustness of the proposed model are discussed.

Increased running speed and axle weight in the transportation network lead to significant dynamic interactions between the vehicles and bridges. It is essential to capture these interactions in fatigue analysis of steel bridges. This paper presents a framework for fatigue evaluation of critical steel bridge details through multi-scale dynamic analysis of the train-track-bridge system and linear elastic fracture mechanics. The multi-scale coupled dynamic analysis allows accurate and efficient computation of fatigue stresses produced by the moving trains in structural details based on a vehicle-bridge analysis model composed of a 3D vehicle model, multi-scale bridge finite element model including the track system, and a wheel–rail interaction model. Field data from an existing steel-truss railway bridge are used to validate the multi-scale analysis method. Enhanced fatigue evaluation of the bridge detail is performed using the computed fatigue load effects and linear elastic fracture mechanics. The effects of the track irregularity and operating train speed on fatigue crack propagation life are investigated. The presented framework is general and can be applied to other types of steel bridges such as the steel-box girder bridge with orthotropic decks.

This paper evaluates the fatigue stress and damage accumulation in steel railway bridges using a dynamic coupled train-bridge system model. The system consists of a spatial vehicle model, a three-dimensional (3D) bridge model, and a wheel-rail interaction model including the effects of track irregularities. The proposed approach was validated by comparing the fatigue stresses of an existing steel railway bridge to those obtained from field structural-health-monitoring data. Fatigue evaluation of the critical structural detail in the bridge was performed on the basis of stress data resulting from the proposed approach and two conventional methods (i.e., the static-influence-line method and the moving-load method). In addition, an investigation on the effects of train speed, track irregularities, and lateral vibrations on fatigue damage and remaining service life was performed. The results show that all three considered factors can significantly affect the fatigue behavior of critical members in railway bridges.

The dynamic responses of the high-speed railway bridge under the train passage can greatly affect the safety of the entire high-speed train and bridge system. Traditionally, these responses are obtained using either field measurement or numerical analysis. Both tools have their own limitations. For instance, the coupling dynamic train-bridge analysis is generally complicated and time-consuming. This paper proposes a novel machine learning based approach for efficient safety evaluation of the high speed train and short span bridge system. Artificial neural networks are established to map the complicated train-bridge system and to attain the critical bridge displacements. The proposed approach incorporates a complete numerical train-bridge system model to produce reliable data for the neural network training, considering multiple significant random features in the train-bridge system. Various neural network architectures are investigated and compared to find optimal ones that have considerable potentials in realizing online response prediction and safety evaluation. Although the proposed approach focuses on the high-speed train and short span bridge, the methodology is general and can also be applied to other scenarios associated with the vehicle-bridge systems.