| 基于强化学习的在线订单配送时隙运能分配 |
| |
| 引用本文： | 陈淮莉,吴梦姣.基于强化学习的在线订单配送时隙运能分配[J].上海海事大学学报,2017,38(2):51-55. |
| |
| 作者姓名： | 陈淮莉 吴梦姣 |
| |
| 作者单位： | 上海海事大学物流科学与工程研究院,上海海事大学物流科学与工程研究院 |
| |
| 基金项目： | 国家社会科学基金(15BGL084)；上海市科学技术委员会科研计划（14DZ2280200）；上海市哲学社会规划课题（2014BGL018） |
| |
| 摘 要： | 为解决在线订单配送效率低、时隙运能分配不均衡和顾客满意度不高的问题,考虑价格和交付期对消费者选择行为的影响建立Logit模型,采用强化学习结合时隙运能分配特点对到达的订单群进行运能分配.算例模拟结果证明:采用强化学习能使每个时隙每辆车的运能分配均衡,且分配方法符合消费者的行为偏好;消费者对时隙价格偏好程度越高商家收益就越低.结论验证了采用强化学习解决时隙运能分配问题的可行性和有效性.
|
| 关 键 词： | 时隙 运能配置 Logit模型 强化学习 |
| 收稿时间： | 2016/11/15 0:00:00 |
| 修稿时间： | 2017/3/9 0:00:00 |
Capacity allocation of delivery time slot for online orders based on reinforcement learning |
| |
| Institution: | Logistics Research Center, Shanghai Maritime University and Logistics Research Center, Shanghai Maritime University |
| |
| Abstract: | In order to solve the lower efficiency of online order delivery, the unbalanced capacity allocation of time slots and the lower customer satisfaction, the Logit model is established considering the influence of the price and lead time on the selection behavior of consumers. Considering the character of capacity allocation of time slot, the orders are assigned to the vehicles by the reinforcement learning. The example simulation results show that: the capacity of every time slot and every vehicle can be balanced by the reinforcement learning and the allocation method accords with the behavioral preference of consumers; the more attention consumers take to the price of time slot, the lower profit retails can get. The conclusion verifies feasibility and effectiveness of adopting the reinforcement learning to solve the capacity allocation of time slot. |
| |
| Keywords: | |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《上海海事大学学报》浏览原始摘要信息 |
| 点击此处可从《上海海事大学学报》下载免费的PDF全文 |
