Abstract:

In order to handle the vehicle scheduling problem caused by the emergence of random customer demands in the logistics distribution process, this paper generated the emergence probabilities, locations, and demands of dummy customers by employing the aggregation and prediction methods with past random customer demand information and experiences empirical data. Based on the analysis of the function of customer dissatisfaction, the objective of minimization of weighting generalized total delivery distribution cost was proposed with the consideration of two factors of the economy of vehicle scheduling plan and customers’ satisfaction. A vehicle scheduling model was established in accordance with the principle of real customer first and dummy customer next. The revised genetic algorithm with local search was designed. The numerical example of Solomon Standard Test verifies the effectiveness and applicability of the proposed model and algorithm. The results show that compared with other existing methods, the vehicle scheduling plan proposed in this paper not only effectively reduces the total delivery cost of logistics company, but also quickly responses to customers demands, which improves the customers satisfaction and the service level.

Key words: vehicle scheduling problem, aggregation and prediction, random customer demand, dummy customer, revised genetic algorithm

摘要： 针对物流配送过程中客户需求随机出现的车辆调度问题，结合客户随机需求信息和经验数据，运用汇集预测方法，生成合理的虚拟客户出现的概率及其位置和需求量；在分析客户不满意度函数的基础上，综合考虑配送车辆调度方案的经济性和客户满意度两方面因素，提出了加权广义配送总费用最小化目标；依照车辆先真实后虚拟客户配送的原则，建立了配送车辆调度模型，设计了与局部搜索相结合的改进遗传算法。Solomon标准测试算例验证了模型与算法的有效性与适用性，结果表明，相比于现有的其他方法，本文方法制定的配送车辆调度方案不仅能有效降低物流企业的总配送成本，而且能快速响应客户需求，提高客户满意度和服务水平。

关键词: 车辆调度问题, 汇集预测, 随机客户需求, 虚拟客户, 改进遗传算法