多因素不确定情况下路面最优养护维修策略决策方法研究

项目名称： 多因素不确定情况下路面最优养护维修策略决策方法研究

项目编号： No.51308335

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 建筑科学

项目作者： 柏强

作者单位： 长安大学

项目金额： 23万元

中文摘要： 合理的路面养护维修策略可以经济有效地保持良好的路面状况并延长路面的使用寿命。路面养护维修策略的制定需要考虑多个因素，例如路面衰坏速度和材料成本等。现实中这些因素多具有不确定性。目前大多数研究都没有充分考虑这些因素不确定性，从而导致获得的路面养护维修策略并非最优。因此，有必要建立一套充分考虑各种因素的不确定性的最优路面维修养护策略的决策方法。本研究首先分析路面养护维修决策中存在的各种具有不确定性的因素并将它们的不确定性量化。然后应用贝叶斯理论（Bayes Theorem）建立在历史路面状况数据不完全的情况下的路面状况评价模型。通过路面生命周期成本和效益的分析， 建立路面养护维修优化问题的数学规划模型。 在求解数学规划模型过程中，采用Monte Carlo Simulation模拟各因素的不确定性，然后设计优化算法求解数据规划模型并开发相应的软件程序。最后用实例验证本研究所提出的理论方法。

中文关键词： 不确定性；路面维修策略；随机优化；生命周期成本；

英文摘要： Pavement maintenance and rehabilitation (M&R) treatments can extend pavement service life and maintain pavements in good condition to provide acceptable level of service to users. The development of a cost-effective pavement M&R scheduling depends on several factors, such as the performance jumps of various M&R treatments, pavement condition deterioration rate, and each M&R treatment's cost. In practice, these factors are associated with uncertainties. Most existing studies on pavement M&R scheduling were based on the deterministic situation and did not consider such uncertainties. As a result, the obtained pavement M&R scheduling is not real optimal in practice. Therefore, a need exists to develop a methodology that can provide optimal pavement M&R scheduling under uncertainty. This study first identifies all possible factors associated with uncertainty and then quantifies their uncertainties. Next, Bayes' theorem is used to establish the pavement performance model under the situation where there are only limited historical pavement condition data. Using life-cycle cost analysis, optimization formulation is established to describe the pavement M&R scheduling problem. Monte Carlo simulation is used to incorporate the uncertainties into the optimization process. Algorithm is then developed to solve the optimizat

英文关键词： Uncertainty；Pavement Maintenance and Rehabilitation Scheduling；Stochastic Programming；Life-Cycle Cost Analysis；