Author(s)

Huiyang Jiang

Corresponding Author

Huiyang Jiang

Abstract

Aiming at the limitations of traditional bridge detection in time and space coverage, data consistency and response speed, this paper proposes a method of municipal bridge health monitoring and risk assessment based on multi-source data fusion. Firstly, a heterogeneous sensor network covering strain, deflection, temperature and dynamic response is constructed, and the dynamic correlation equation of strain-temperature-deflection is established by introducing Cauchy-hoff plate theory, so as to realize the mechanism-level fusion of local measurement to the whole deformation field. Then, the stiffness reduction coefficient is taken as the performance state parameter, and Bayesian updating is used to fuse the prior design information and real-time monitoring data, which significantly reduces the cognitive uncertainty under small samples. Finally, the time-varying failure probability is calculated by Monte Carlo simulation, and the risk evolution curve is drawn to realize the transition from post-disposal to early warning. The example of 30 m prestressed concrete simply supported beam bridge shows that the error between strain inversion deflection and measured peak value is less than 5%, and the stiffness uncertainty is reduced by 70%. The system can issue intermediate warning four months before the risk accelerates, which provides reliable decision support for the accurate and intelligent management and maintenance of municipal bridges.

Keywords

Municipal Bridge, Health Monitoring, Risk Assessment, Multi-source Data Fusion