Dynamic collaborative scheduling for multi-depot fresh food delivery
To address the vehicle scheduling problem in fresh instant delivery with multiple distribution centers under limited vehicle capacity and dynamic demand, this study proposes a collaborative capacity allocation-based vehicle scheduling method. A rolling-horizon strategy is adopted, and a multi-stage vehicle scheduling model is developed to minimize the total scheduling cost. A two-layer hybrid optimization algorithm incorporating order spatiotemporal feature perception is designed. In the upper layer, an evolutionary mechanism optimizes delivery task assignment, while the lower layer optimizes vehicle routes under given task assignments. For routing problems with time-window and time-dependent speed characteristics, proximal policy optimization and a 3-opt correction operator optimize vehicle-level order visiting sequences, forming a complementary optimization mechanism with adaptive large neighborhood search. Multi-scale case studies show that collaborative capacity allocation reduces delivery costs and alleviates the mismatch between dynamic demand and limited vehicle capacity, while the proposed algorithm improves optimization performance and stability.
Authors
- Xiaoting Huang (ORCID: https://orcid.org/0000-0001-9596-2971)
- Shi Chuwei
- Changxi Ma (ORCID: https://orcid.org/0000-0002-0250-5462)
Institutions
- Lanzhou Jiaotong University (CN)
- Lanzhou University of Finance and Economics (CN)
Publication Details
- Journal
- Transportation Letters
- Published
- 2026-11-18
- DOI
- https://doi.org/10.1080/19427867.2026.2748867
- Primary Topic
- Vehicle Routing Optimization Methods
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Natural Science Foundation of China