摘要
为解决电商订单取消行为导致的配送资源浪费问题,根据消费者对时隙价格的偏好建立Logit模型,通过机会成本确立时隙运能超售点,采用强化学习结合时隙运能分配特点研究订单取消行为下的时隙运能超售策略。模拟结果证明:强化学习能使订单运能均匀分配;相比传统方式,采用超售策略能够提高总收益,在最优超售点下总收益最大。利用该模型可得出不同订单取消率下的最优超售点以及收益增加率,为商家制定相关销售和运输策略提供参考。
To solve the waste problem of distribution resource caused by order cancellation behavior of e-commerce,the Logit model is established according to the consumer preference for time slot prices,the oversold point of time slot delivery capacity is determined through opportunity cost,and the oversold strategy of time slot delivery capacity under order cancellation behavior is studied by reinforcement learning combined with the characteristics of time slot delivery capacity allocation.The simulation results show that:the reinforcement learning can make the order delivery capacity distribution even;compared to the traditional method,the oversold strategy can increase the total revenue,and the total revenue can be maximized at the optimal oversold point.The optimal oversold points and the revenue increase rates under different order cancellation rates are obtained by the model,which provides reference for merchants to formulate relevant sales and transportation strategy.
作者
刘鹏宇
陈淮莉
LIU Pengyu;CHEN Huaili(Institute of Logistics Science & Engineering, Shanghai Maritime University, Shanghai 201306, China)
出处
《上海海事大学学报》
北大核心
2018年第4期38-43,共6页
Journal of Shanghai Maritime University
基金
上海市科学技术委员会重点项目(16040501800)
关键词
时隙运能
LOGIT模型
强化学习
超售策略
time slot delivery capacity
Logit model
reinforcement learning
oversold strategy