ABSTRACT

This paper introduces the Fleet Size and Mix Vehicle Routing Problem with Synchronized Visits (FSM-VRPS), an extension of the Vehicle Routing Problem with Synchronization (VRPS), where a mixed fleet composed of electric and conventional bikes, and passenger cars having different acquisition costs are considered. The problem consists of planning a set of different vehicle routes to serve a set of clients who may require more than one visit by different healthcare specialists, and some of these visits should be synchronized. Moreover, each client must be visited within a specified time window. In addition, the problem uses bikes to reduce Carbon Dioxide (CO₂) emission for environmentally cleaner routing operations. This problem has many real-life applications, such as the scheduling of visits for homecare givers in the healthcare sector. We present a mixed integer linear-programming formulation and develop a Multi-Start Adaptive Large Neighborhood Search with Threshold Accepting algorithm. The results showed that our algorithm is highly effective on the FSM-VRPS, as well as on the heterogeneous VRPS. We also demonstrate the advantage of adopting different types of vehicles in terms of reducing the number of vehicles and costs. The analysis of the results also indicated that the new components added to the standard Adaptive Large Neighborhood Search algorithm enhanced intensification and diversification mechanisms during the search process.

摘要

本文介绍了具有同步访问的车队规模和混合车辆路线问题 (FSM-VRPS)，这是同步访问车辆路线问题 (VRPS) 的扩展，其中混合车队由电动自行车和传统自行车以及具有不同获取方式的乘用车组成成本考虑。该问题包括规划一组不同的车辆路线，以服务于可能需要不同医疗保健专家多次访问的一组客户，并且其中一些访问应该同步。此外，必须在指定的时间窗口内访问每个客户端。此外，该问题使用自行车来减少二氧化碳（CO ₂) 排放，用于环境更清洁的路由操作。这个问题有许多现实生活中的应用，例如医疗保健部门的家庭护理人员的就诊安排。我们提出了一个混合整数线性规划公式，并开发了一种具有阈值接受算法的多起点自适应大邻域搜索。结果表明，我们的算法在 FSM-VRPS 以及异构 VRPS 上都非常有效。我们还展示了采用不同类型车辆在减少车辆数量和成本方面的优势。对结果的分析还表明，标准自适应大邻域搜索算法中添加的新组件增强了搜索过程中的强化和多样化机制。