Multicriteria Optimization Method for Network-Level Bridge Management

A multicriteria optimization method for analyzing important capital investment decisions involved in managing bridge infrastructure is presented. The condition assessment and decision variables of the method can be adapted to analyze a population of small and medium-size bridges or a population of spans of a large bridge. Condition ratings of various bridge structural elements are used to assess the condition needs of four major components. Subsequent use of this information leads to characterization of bridge condition by defining bridge states. State increment models are used to identify suitable treatment options for each state and predict the variable time over which state increments (or transitions) occur. These state increment models are incorporated into an optimization method that has three major steps: (a) identification of objective functions representing the multiple decision criteria, (b) assessment of the importance of each objective in achieving the numerical goals targeted by decision makers, and (c) formulation of a goal programming model. The goal program determines an optimal multi-year bridge program that minimizes the weighted sum of deviations from goals. Important results from the analysis of capital program scenarios for more than 800 small and medium-size bridges managed by the New York State Thruway Authority are presented. It is concluded that the multicriteria optimization method provides a useful tool to analyze multiple goal-oriented scenarios for a bridge capital program and establish a relationship between average network condition rating and total expenditure.