Optimal Sensor Placement Strategy Considering Bayesian Identification Uncertainty and State Change in Bridge Health Monitoring

In structural health monitoring (SHM) projects, sensor placement is critically linked to the uncertainty of identification outcomes, directly determining the reliability of monitoring results. Although existing studies predominantly focus on the relationship between sensor configurations and uncertainties in modal parameter identification, the influence of actual structural changes on identification uncertainty remains underexplored. This paper proposes an adaptive sensor placement strategy for bridge SHM systems, tailored to accommodate varying structural health states. Leveraging a Bayesian two-stage system identification framework, optimal sensor layouts were independently determined for each phase with distinct objectives and subsequently integrated. During the modal identification stage, optimal sensor positions were selected by minimizing modal parameter uncertainty to ensure robust identification. In the subsequent phase, sensor locations were dynamically adjusted to maximize sensitivity to structural changes through quantification of discrepancies between measured parameters and analytical models. A composite objective function was formulated to harmonize these dual priorities. Numerical simulations on a simply supported beam demonstrate that the proposed sensor arrangement concurrently reduces parameter identification uncertainty and preserves sensitivity to structural variations. Field validation on a 60-m continuous girder bridge confirms the method’s practical efficacy. The strategy is applicable to both aging structures (with localized damage or repair histories) and newly constructed facilities. Notably, when integrated with environmental vibration testing techniques, it enables operational-phase sensor reconfiguration to target emerging damage zones, thereby ensuring sustained SHM system reliability and prioritized monitoring of critical regions during long-term service.