| Summary: | The main objective of a bridge manager is to find the best maintenance plan for a group of bridges over a prescribed time horizon. The bridge manager usually faces conflicting ob- jectives, as nlaintenance plans resulting in safer and less deteriorated structures also lead to higher costs. In general, the problem is posed as a deterministic single-objective optimiza- tion where cost is minimized keeping performance above pre-defined thresholds. However, single-objective optimization results in only one optimal solution that does not provide the advantages or disadvantages of considering other objectives and constraints. In addition, the effects of uncertainties are not taken into account or are included in a very simplified way. The bridge manager obtains, in this way, only one deterministic optimum maintenance plan, and not a set of different probabilistic maintenance solutions from which the best, for each particular situation, can be chosen. In this paper, a full probabilistic multi-objective approach to bridge maintenance con- sidering single maintenance types is developed. This approach is based on the latest de- velopments in bridge management by considering probabilistic continuous performance in- dicators and probabilistically defined objectives and constraints. The problem is solved by using multi-ob jective genetic algorithms and Latin hypercube sampling technique. Multi- objective applications to existing reinforced concrete bridge components under probabilistic deterioration and probabilistic-defined single maintenance types are presented and discussed.
|
|---|