This study develops a new integer-programming model to address the network-wide daily road pavement rehabilitation scheduling problem. In the model, the crew organization and work zone schedule are jointly optimized daily, with the objective of minimizing both the operational cost and user travel time. A day-to-day traffic dynamics model is applied to capture the non-equilibrium traffic evolution against network supply variation over the planning horizon, which leads to a simulation-based optimization problem. To solve this challenging problem, a two-stage hybrid heuristic solution method is proposed. In the first stage, hybrid tabu search (TS) meta-heuristics are comparatively developed to identify a group of active crew work routes without time slacks. The obtained crew routes are then fed to the second stage for work zone scheduling via a discrete compass search algorithm. Some important findings are obtained from numerical experiments. First, crew routing (or crew organization) is the dominant decision in the studied problem, and a desirable work zone schedule encourages a crew to execute the assigned tasks continually. The findings can be used to develop simplified and efficient solution algorithms. Second, the hybrid TS meta-heuristics developed for crew routing exhibit superior performance compared to other solution methods. Finally, a well-defined model for the current problem should consider both user travel time and operation cost. Our model enables decision-makers to make an effective trade-off between these two objectives. An effective measure is suggested to evaluate the cost-effectiveness of budget investment decisions when budgets are limited.

中文翻译：

---

综合人员组织和工作区调度，实现全网日常道路路面修复


本研究开发了一种新的整数规划模型来解决网络范围内的日常​​道路路面修复调度问题。在该模型中，每天联合优化船员组织和工作区时间表，以最大限度地减少运营成本和用户出行时间。应用日常流量动态模型来捕获规划范围内网络供应变化的非平衡流量演变，从而产生基于模拟的优化问题。为了解决这个具有挑战性的问题，提出了一种两阶段混合启发式求解方法。在第一阶段，混合禁忌搜索（TS）元启发法相对发达，用于识别一组没有时间松弛的活跃船员工作路线。然后，获得的乘员路线将通过离散罗盘搜索算法输入第二阶段，用于工作区调度。一些重要的发现是从数值实验中获得的。首先，机组路线（或机组组织）是所研究问题的主导决策，理想的工作区时间表鼓励机组持续执行分配的任务。研究结果可用于开发简化且高效的解决算法。其次，与其他解决方案方法相比，为机组路由开发的混合 TS 元启发法表现出卓越的性能。最后，针对当前问题的明确定义的模型应同时考虑用户出行时间和运营成本。我们的模型使决策者能够在这两个目标之间进行有效的权衡。提出了在预算有限的情况下评估预算投资决策成本效益的有效措施。