基于DRL与元启发式算法的冷链物流路径优化研究
Research on Cold Chain Logistics Route Optimization Based on DRL and Meta-Heuristic Algorithm
摘要: 随着冷链物流市场的快速扩容与智能化升级，多约束路径优化问题成为行业发展的关键瓶颈。深度强化学习(Deep Reinforce Learning, DRL)与元启发式算法相融合的方法将二者优势结合起来，在路径优化问题中展现巨大潜力。据此，本文分析了不同融合模式的设计逻辑差异，探讨了各类方法的优势与局限性，并展望了未来研究方向。
Abstract: With the rapid expansion and intelligent upgrading of the cold chain logistics market, the problem of multi-constraint path optimization has become a key bottleneck in the development of the industry. The fusion of deep reinforcement learning (DRL) and meta-heuristic algorithms combines the advantages of the two and shows great potential in path optimization problems. Based on this, this paper analyzes the differences in design logic among various fusion models, discusses the advantages and limitations of various methods, and looks forward to future research directions.
文章引用：赵墨然, 周艳聪, 周钰京, 冯大中, 廖欧珣, 洪宇恒. 基于DRL与元启发式算法的冷链物流路径优化研究[J]. 管理科学与工程, 2025, 14(6): 1020-1024. https://doi.org/10.12677/mse.2025.146119

参考文献

[1] 李卫朝. 基于深度强化学习的动态多车型绿色车辆路径优化研究[D]: [硕士学位论文]. 西安: 西安电子科技大学, 2022.
[2] 方文婷, 艾时钟, 王晴, 等. 基于混合蚁群算法的冷链物流配送路径优化研究[J]. 中国管理科学, 2019, 27(11): 107-115.
[3] 黄继磊. 基于粒子群自进化的冷链物流运输路径优化方法[J]. 包装工程, 2025, 46(11): 285-293.
[4] 张天瑞, 祝芳芳, 牛慧媛. 基于改进哈里斯鹰算法的生鲜品冷链物流路径优化研究[J]. 重庆师范大学学报(自然科学版), 2025, 42(2): 1-13.
[5] 江云倩, 杨慧敏, 彭程, 等. 考虑碳排放和时间窗的冷链物流配送路径优化研究[J]. 包装工程, 2024, 45(3): 262-268.
[6] 李军涛, 刘明月, 刘朋飞. 生鲜农产品多车型冷链物流车辆路径优化[J]. 中国农业大学学报, 2021, 26(7): 115-123.
[7] 吴暖, 代焕杰, 李季涛, 等. 考虑时间容忍度的冷链物流配送路径多目标优化[J]. 交通运输系统工程与信息, 2023, 23(2): 275-284.
[8] Kool, W., van Hoof, H. and Welling, M. (2018) Attention, Learn to Solve Routing Problems! arXiv: 1803.08475.
[9] Ma, Y., Liu, Q. and Tang, J. (2023) Learning to Search Feasible and Infeasible Regions of Routing Problems with Flexible Neural k-Opt. arXiv: 2310.18264.
[10] 王万良, 陈浩立, 李国庆, 等. 基于深度强化学习的多配送中心车辆路径规划[J]. 控制与决策, 2022, 37(8): 2101-2109.
[11] Silva, M.A.L., Souza, S.R.de, Souza, M.J.F. and Bazzan, A.L.C. (2019) A Reinforcement Learning-Based Multi-Agent Framework Applied for Solving Routing and Scheduling Problems. Expert Systems with Applications, 131, 148-171.
[12] Ye, H., Wang, J., Cao, Z., Liang, H. and Li, Y. (2023) DeepACO: Neural-Enhanced Ant Systems for Combinatorial Optimization. arXiv: 2309.14032.
[13] Rodríguez-Esparza, E., Masegosa, A.D., Oliva, D. and Onieva, E. (2024) A New Hyper-Heuristic Based on Adaptive Simulated Annealing and Reinforcement Learning for the Capacitated Electric Vehicle Routing Problem. Expert Systems with Applications, 252, Article ID: 124197. [Google Scholar] [CrossRef] 
[14] 许波, 姜艺, 刘玉敏, 等. 用于冷链食品运输过程的结合Q学习机制的变邻域蚁群优化调度方法[P]. 中国专利, CN115063177B. 2024-03-08.
[15] Leng, L., Jin, Q., Chen, T., Wan, A. and Wang, Z. (2024) Energy-Conserving Cold Chain with Ambient Temperature, Path Flexibility, and Hybrid Fleet: Formulation and Heuristic Approach. International Journal of Production Research, 63, 26-50. [Google Scholar] [CrossRef] 
[16] Wu, R., Wang, R., Hao, J., Wu, Q., Wang, P. and Niyato, D. (2025) Multiobjective Vehicle Routing Optimization with Time Windows: A Hybrid Approach Using Deep Reinforcement Learning and Nsga-ii. IEEE Transactions on Intelligent Transportation Systems, 26, 4032-4047. [Google Scholar] [CrossRef] 
[17] Ma, Y., Li, J., Cao, Z., Song, W., Zhang, L., Chen, Z. and Tang, J. (2021) Learning to Iteratively Solve Routing Problems with Dual-Aspect Collaborative Transformer. Advances in Neural Information Processing Systems, 34, 11096-11107.
[18] Zhao, J., Mao, M., Zhao, X. and Zou, J. (2021) A Hybrid of Deep Reinforcement Learning and Local Search for the Vehicle Routing Problems. IEEE Transactions on Intelligent Transportation Systems, 22, 7208-7218.
[19] 张景玲, 冯勤炳, 赵燕伟, 等. 基于强化学习的超启发算法求解有容量车辆路径问题[J]. 计算机集成制造系统, 2020, 26(4): 1118-1129.
[20] Li, P., Zheng, Y., Tang, H., Fu, X. and Hao, J. (2024) EvoRainbow: Combining Improvements in Evolutionary Reinforcement Learning for Policy Search. Proceedings of the 41st International Conference on Machine Learning, 235, 29427-29447.

为你推荐