Incorporating Priority Preferences into Pavement Maintenance Programming

Abstract

Traditionally, it has been a common practice to apply priority weights to selected parameters in the process of optimal programming of pavement maintenance. However, there are several issues associated with this approach of incorporating priority preferences into pavement maintenance programming. For instance, applying priority weights to selected problem parameters will lead to a suboptimal solution with respect to the original objective function (such as minimal total maintenance cost or maximum pavement condition). The decision makers may not be aware of this consequence and the magnitude of loss in optimality caused by their choice of priority scheme. This paper proposes an improved methodology of incorporating priority preferences into pavement maintenance programming to overcome these problems. Instead of applying priority weights directly into the mathematical formulation of maintenance programming, priority preferences are handled in two stages of postprocessing of the optimal programming process, namely, a tie-breaking analysis and a trade-off analysis. The optimal programming problem is first solved without applying priority weights to any parameters of the problem. This ensures that the optimality of the solution is not disturbed. In the tie-breaking postprocessing, prioritized maintenance activities are identified to replace lower-priority activities in the solution without affecting the optimality of the solution. Finally, a trade-off analysis is performed to introduce more prioritized activities into the solution on the basis of the willingness of the highway agency to accept some loss in optimality.