摘要
对物流长途配送路径进行合理规划可以节省运输车辆的数量,缓解交通压力.为解决智慧物流园区长途配送路径规划方法存在规划路径不是最优、规划时间较长的问题,提出基于深度强化学习的智慧物流园区长途配送路径规划方法.首先,将最短配送距离作为规划目标,构建智慧物流园区长途配送路径规划模型;然后,对客户状态信息进行融合处理,完成对注意力机制的构建;最后,利用深度强化学习,对规划模型的路线规划能力迭代训练,将训练后模型输出的结果作为最佳配送路径规划结果.实验结果证明,使用该方法能够有效得出最佳配送路径,路径长度基本为4~8 m,获取最佳平均路径长度对应的时间小于20 ms,规划性能较好,规划时间较短,具有较好的规划效果.
Reasonable planning of logistics long-distance distribution path can save the number of transportation vehicles and ease the traffic pressure.However,the current long-distance distribution path planning method of smart logistics park has the problems that the planning path is not optimal and the planning time is long.Therefore,we propose a long-distance distribution path planning method based on deep reinforcement learning for smart logistics parks.Firstly,the shortest distribution distance is taken as the planning goal to build a long-distance distribution path planning model of the smart logistics park;then,the customer state information is fused and processed to complete the construction of the attention mechanism;finally,the route planning capability of the planning model is iteratively trained using deep reinforcement learning,and the output of the trained model is taken as the best distribution path planning result.The experimental results prove that the best distribution path can be effectively derived using this method,and the path length is basically controlled within the range of 4m~8m,and the time corresponding to obtaining the best average path length is less than 20ms,with better planning performance and shorter planning time,which has better planning effect.
作者
解晓乐
XIE Xiao-le(School of Information Engineering and Media,Hefei Technology College,Hefei Anhui 230013,China)
出处
《广州航海学院学报》
2024年第1期30-34,68,共6页
Journal of Guangzhou Maritime University
基金
2023年度安徽省高校自然科学研究重点项目(2023AH052550)。
关键词
智慧物流园区
路径规划
长途配送
深度强化学习
Smart logistics park
path planning
long-distance distribution
deep reinforcement learning
