车辆合乘匹配问题中服务需求分派算法研究  

Research on service requirement distribution algorithms in carpooling problems

作　　者：邵增珍[1,2] 王洪国[1] 刘弘[1] 孟春华[2,3] 张子鹏[2,3] 

机构地区：[1]山东师范大学信息科学与工程学院,济南250014 [2]山东省物流优化与预测工程技术研究中心,济南250014 [3]山东师范大学管理科学与工程学院,济南250014

出　　处：《清华大学学报（自然科学版）》2013年第2期252-258,264,共8页Journal of Tsinghua University(Science and Technology)

基　　金：国家自然科学基金资助项目(60970004);山东省自然科学基金资助项目(ZR2011FQ029;ZR2011FL026)

摘　　要：在目前交通拥塞及物流成本居高不下的情况下,通过车辆合乘的方式降低车辆空载率、降低物流成本具有重要意义。在多车辆环境下,如何恰当地将服务需求分派到合适的车辆上是需要首先解决的问题。该文对确定性多车辆合乘匹配问题展开研究,提出针对服务需求分派的启发式聚类算法。基于服务需求及车辆特征提出匹配度概念,并基于启发式算法将服务需求分派到具体某一车辆。实际算例验证:该算法在可接受时间范围内,可有效提高搭乘成功率,降低车辆运行成本。The traffic congestion problem and logistics costs are becoming more and more serious at present. Carpooling not only improves vehicle utilization rates, but also reduces logistics costs. In multi vehicle environment, how to distribute service requirements to proper vehicles should first be considered. In this paper, a heuristic clustering algorithm was developed to solve the grouping progress in a deterministic carpooling problem. The concept of matching degree was proposed based on the characteristics of service requirements and vehicles to distribute requirements to a specific vehicle using heuristic algorithm. Simulation results show that the algorithm significantly improves success ride rates and reduces total vehicle costs.

关 键 词：车辆合乘问题 启发式聚类算法 服务需求分派 匹配度 

分 类 号：TP15[自动化与计算机技术—控制理论与控制工程]