Network-Level Multi-objective Optimal Maintenance and Rehabilitation Scheduling

Lu Gao,Chi Xie,Zhanmin Zhang
2010-01-01
Abstract:The focus of this paper is a multiobjective pavement maintenance and rehabilitation scheduling problem, which aims at optimizing the pavement conditions and budget utilization in a simultaneous manner. In general, a multiobjective optimization problem results in a set of Pareto-optimal or efficient solutions instead of a single optimal one. While it is possible to know the objective preference a priori and incorporate this preference information into the solution search process, in most cases, it is more general for a planner to first identify the whole Pareto optimal solution set and then choose one or a few solutions from the solution set for a best system performance trade-off. The multi-objective maintenance and rehabilitation scheduling problem has been addressed by some heuristic procedures, such as the weighting method, constraint method or metaheuristics, which, however, cannot guarantee the complete Pareto optimal solutions set. In this paper, we formulate the multi-objective problem as a linear programming problem and apply a parametric method to exhaust the Pareto-optimal solution set. The proposed method has been successfully implemented in a case study of the Dallas District. Sensitivity analyses with the budget constraint and the condition requirement constraint are also carried out. A comparison is made between the widely used weighting method and the parametric method.
What problem does this paper attempt to address?