Robust Maintenance Planning of Deteriorating Bridges Considering Uncertainty of Condition Index Weights

This study proposes a novel robust maintenance planning framework for deteriorating bridges that explicitly addresses the uncertainty in the weighting of bridge condition indices. The primary objective is to ensure that the bridge condition rating requirements are satisfied through predetermined maintenance interventions, even in the presence of uncertainty in the weights of the bridge condition indices. In addition, variations in the weighting of the bridge condition indices for selected bridges in Chongqing, China, were investigated and analyzed. The observed fluctuations in these weights were incorporated into a maintenance decision-making model, thereby enabling the development of a robust planning approach. Thus, a robust optimization model was formulated to explicitly integrate these uncertainties into the maintenance planning process. This framework enables the formulation of an optimal set of preventive maintenance actions that are strategically designed to minimize the overall maintenance costs while ensuring compliance with the prescribed condition rating thresholds, even under perturbations in index weight values. The effectiveness of the proposed method was validated through a series of numerical examples involving typical prestressed concrete girder bridges as representative case studies. For comparative purposes, conventional (non-robust) multi-objective maintenance plans that do not account for the uncertainty in bridge performance are also constructed and presented. These baseline plans served as benchmarks to highlight the advantages of the robust approach. Results confirm that the proposed robust method can generate feasible maintenance plans under varying disturbances in bridge condition index weights. Importantly, the method is shown to perform effectively in simultaneously ensuring that the prescribed bridge condition ratings are satisfied and maintenance costs are minimized.