摘要: |
Bridge management involves making decisions related to selecting the most cost-effective bridge improvement strategies both at the project and network levels. Typical cost variables include agency costs, user costs, discount rates, inflation, or interest rates, life-cycle costs, etc. Lack of reliable sources for accurate deterministic cost data has been identified as one of the shortcomings of current models of bridge management. Uncertainties can easily lead to making the wrong decisions, especially in selecting the best from pairs of closely ranked competing strategies. Historical records can be used to generate probabilistic estimates. Expert opinions may also be used to suggest subjective estimates or used to complement estimates obtained from historical records. This paper presents an overview of this uncertainty problem in cost estimating of bridge decision variables and also discusses suggestions on how to handle the uncertainty using analytical tools such as the fuzzy sets and probability theories. The decision-making algorithms in some existing bridge management systems and cost models are used to illustrate the suggested uncertainty-handling methods and also show how the uncertainties can affect the overall decision. |