Answers are grounded in this paper and cite their sources.
This paper proposes a physics-informed neural operator framework, termed physics-informed neural operator-vehicle-bridge interaction (PINO-VBI), for efficiently predicting the dynamic responses of vehicle-bridge interaction (VBI) systems. The numerical validation focuses on controlled low-speed VBI scenarios, in which the vehicle speed ranges from 4 to 8 m/s. The model integrates a branched Fourier neural operator (BFNO) with physics-informed loss terms to accommodate the frequency heterogeneity among vehicle subsystems and incorporate governing differential equations during training. To address the disparity in response characteristics across subsystems, a dual-branch BFNO architecture is developed, allowing flexible allocation of Fourier modes and hyperparameters for different frequency components. The network is trained using synthetic datasets generated from VBI simulations, incorporating variations in vehicle speed, system parameters, and random road excitations. A comprehensive hyperparameter study is conducted to assess the sensitivity of prediction performance to model architecture. Furthermore, the impact of embedding different types of physical constraints, including equation residuals, velocity constraints, and acceleration constraints, is systematically analyzed. Comparative experiments demonstrate that the proposed PINO-VBI outperforms standard FNO-based PINO models, particularly in accurately capturing velocity and acceleration responses critical to ride comfort and structural safety assessment. This study highlights the advantages of combining physics-informed learning with frequency-aware neural operators in modeling VBI systems, and provides a methodological basis for efficient surrogate simulation and future digital twin applications in bridge health monitoring and dynamic safety evaluation.
No takes yet. Share an insight, caveat, or question.
Zhu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: