研究期間：10108~10207;The main objectives of the research include the development of solution methods and the comparison for different maintenance strategies in transportation infrastructure life-cycle management. First, most of the the existing infrastructure performance models in the literature assumed a single deterioration mode. However, many studies showed that facility deterioration trends might depend on maintenance history. Moreover, the trends could also be affected by the selected maintenance actions. Therefore, this research plans to adopt hybrid dynamic models (HDM) to simulate the life cycles of transportation infrastructure. HDMs are capable of incorporating multiple deterioration modes and switching between them automatically. As a result, they are appropriate for the life-cycle optimization using multiple actions with heterogeneous effects. However, these models are complicated and require significant computational time to optimize for large-scale problems (a large number of facilities or long planning periods). Thus, the first major task of this research is to develop solution methods for solving this type of problems. Threshold-based maintenance and pre-determined plan maintenance are the most widely used strategies in transportation life-cycle management. The pre-determined plan strategy generates detailed plans that include the timing and type of actions that should be applied at each time period. On the other hand, threshold-based maintenance means that facilities receive maintenance when their performances reache the thresholds. The two strategies have both advantages and disadvantages. Therefore, the second major task in this research is to use the hybrid dynamic models and the solution methods to solve, analyze, and compare the above two strategies. The goal is to propose guidelines for selecting maintenance strategies in different scenarios such as budget availability and problem scales.
