Safety-based multi-objective life-cycle management system for steel box girder bridges

In many countries including Korea, a number of studies have been carried out to propose methods for managing the deterioration of bridges in order to obtain the optimum maintenance strategy associated with higher performance and lower cost. A number of computer-aided bridge management systems have been developed based on these methods. Bridge condition based on visual inspection was monitored and used as a threshold. Because of the limitation of visual inspection as a safety index, other safety related indexes such as a damage index or reliability index have been adopted while the bridge management system has evolved as a decision making tool equipped with the capability of generating the optimal maintenance scenario. However, most bridge management systems still have some restrictions in producing feasible solutions. Some important restrictions are related to the multi-objectives, cost-maintenance interaction and subordinate relation between bridge members. In this study, a new bridge management method is proposed and computer software based on this method has been developed to obtain the life-cycle optimum maintenance strategy for deteriorating steel box girder bridges. A multi-objective optimisation problem is formulated and the Genetic Algorithm is applied to obtain optimal tradeoff maintenance scenarios.