Abstract:With the rapid development of big data, the Internet of Things (IoT), and other technological advancements, digital twin (DT) technology is increasingly being applied to the field of bridge structural health monitoring. Achieving the precise implementation of DT relies significantly on a dual-drive approach, combining the influence of both physical model-driven and data-driven methodologies. In this paper, two methods are proposed to predict the displacement and dynamic response of structures under strong winds, namely, a Bayesian Neural Network (BNN) model based on Bayesian inference and a finite element model (FEM) method modified based on genetic algorithms (GAs) and multi-objective optimization (MOO) using response surface methodology (RSM). The characteristics of these approaches in predicting the dynamic response of large-span bridges are explored, and a comparative analysis is conducted to evaluate their differences in computational accuracy, efficiency, model complexity, interpretability, and comprehensiveness. The characteristics of the two methods were evaluated using data collected on the Forth Road Bridge (FRB) as an example under unusual weather conditions with strong wind action. This work proposes a dual-driven approach, integrating machine learning and FEM with GNSS and Earth Observation for Structural Health Monitoring (GeoSHM), to bridge the gap in the limited application of dual-driven methods primarily applied for small- and medium-sized bridges to large-span bridge structures. The research results show that the BNN model achieved higher R2 values for predicting the Y and Z displacements (0.9073 and 0.7969, respectively) compared to the FEM model (0.6167 and 0.6283). The BNN model exhibited significantly faster computation, taking only 20 s, while the FEM model required 5 h. However, the physical model provided higher interpretability and the ability to predict the dynamic response of the entire structure. These findings help to promote the further integration of these two approaches to obtain an accurate and comprehensive dual-driven approach for predicting the structural dynamic response of large-span bridge structures affected by strong wind loading.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: how to accurately predict the dynamic response of long - span bridges under strong winds. Specifically, the research aims to improve the health monitoring ability of long - span bridge structures under extreme weather conditions by combining physics - model - driven and data - driven methods. ### Research Background and Problems 1. **Vulnerability of Long - Span Bridges** - Long - span bridges (especially suspension bridges) are prone to large structural displacements and local damage under extreme weather conditions such as strong winds, which affect their stability and service performance. - Therefore, accurately predicting the dynamic response of long - span bridges under extreme weather conditions is crucial for the health monitoring of bridge structures. 2. **Limitations of Existing Methods** - Although traditional finite element analysis (FEM) can simulate the dynamic response of bridge structures, it has problems such as high computational cost and high model complexity in practical applications. - Data - driven methods (such as machine learning) have strong non - linear fitting capabilities, but are sensitive to noise and have poor interpretability. 3. **Research Objectives** - Propose a dual - driven method combining the finite element model and Bayesian neural network (BNN) to more accurately predict the dynamic response of long - span bridges under strong winds. - Evaluate the application potential in the health monitoring of long - span bridge structures by comparing the computational accuracy, efficiency, model complexity, interpretability and comprehensiveness of these two methods. ### Method Overview 1. **Finite Element Model Based on Response Surface Modification (RSM - FEM)** - Use genetic algorithm (GA) and multi - objective optimization (MOO) to modify the finite element model and improve the accuracy of the model. - Reduce the computational cost through the response surface method (RSM) to achieve efficient optimization. 2. **Data - Driven Method Based on Bayesian Neural Network (BNN)** - Use Bayesian inference to update the posterior distribution of weights and biases, providing better anti - noise ability and generalization performance. - It has strong fast - learning ability, especially suitable for small - sample situations. ### Experimental Verification - Take the Forth Road Bridge (FRB) as an example and use the actual wind speed data collected by GNSS and GeoSHM for verification. - Compare the performance of the two methods in predicting displacements in the Y - direction and Z - direction. The results show that the BNN model is superior to the FEM model in terms of calculation speed and accuracy. ### Conclusions and Prospects - This research proposes a dual - driven method combining physical models and data - driven methods, providing new ideas for predicting the dynamic response of long - span bridge structures under strong winds. - Future research will further integrate these two methods to achieve more accurate and comprehensive bridge structure health monitoring. Through this method, researchers hope to promote the application of digital twin (DT) technology in the field of bridge engineering and improve the safety and reliability of long - span bridges.

Application of Response Surface-Corrected Finite Element Model and Bayesian Neural Networks to Predict the Dynamic Response of Forth Road Bridges under Strong Winds

Examination of load-deformation characteristics of long-span bridges in harsh natural environments based on real-time updating artificial neural network

Dynamic response prediction of vehicle-bridge interaction system using feedforward neural network and deep long short-term memory network

Numerical Investigation Of The Dynamic Responses Of Long-Span Bridges With Consideration Of The Random Traffic Flow Based On The Intelligent Aco-Bpnn Model

Prediction Model for Long-Term Bridge Bearing Displacement Using Artificial Neural Network and Bayesian Optimization

Prediction Method of Bridge Static Deformation Based on Dynamic Test

Nonlinear Unsteady Aerodynamic Forces Prediction and Aeroelastic Analysis of Wind-Induced Bridge Response at Multiple Wind Speeds: A Deep Learning-Based Reduced-Order Model

Mapping Model Between Lateral Responses of Bridges under Random Traffic Loads Based on Data-Driven Approach

Probabilistic Framework with Bayesian Optimization for Predicting Typhoon-Induced Dynamic Responses of a Long-Span Bridge

Performance Prediction for Steel Bridges Using SHM Data and Bayesian Dynamic Regression Linear Model: A Novel Approach

A sequence-to-sequence model for joint bridge response forecasting

Physically Guided Estimation of Vehicle Loading-Induced Low-Frequency Bridge Responses with BP-ANN

Deep Learning-Based Prediction of Wind-Induced Lateral Displacement Response of Suspension Bridge Decks for Structural Health Monitoring

Real-Time Bridge Deflection Prediction Based on a Novel Bayesian Dynamic Difference Model and Nonstationary Data

Prediction of bridge structure deformation and strain based on dynamic testing and intelligent algorithms

High-Efficiency Finite Element Model Updating of Bridge Structure Using a Novel Physics-Guided Neural Network

Bayesian dynamic noise model for online bridge deflection prediction considering stochastic modeling error

Bayesian Finite Element Model Updating and Assessment of Cable-Stayed Bridges Using Wireless Sensor Data

High-performance beam finite element for predictive response in monitoring existing bridges

Digital Modeling on the Nonlinear Mapping Between Multi‐source Monitoring Data of In‐service Bridges

Machine learning‐aided prediction of windstorm‐induced vibration responses of long‐span suspension bridges