An investigation into multi-objective decision-making in fresh cold chain supply chain networks within a dual distribution framework
Abstract Aiming at the problems of facility location, path planning, and flow allocation in a cold chain supply network that incorporates both direct and transit (traditional two-stage) distribution modes, this paper introduced different types of cold storage and different types of cold-chain trucks...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2025-07-01
|
| Series: | Complex & Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s40747-025-02021-5 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Aiming at the problems of facility location, path planning, and flow allocation in a cold chain supply network that incorporates both direct and transit (traditional two-stage) distribution modes, this paper introduced different types of cold storage and different types of cold-chain trucks, and constructed a multi-mode coordinated distribution system to adapt to the comprehensive requirements of fresh agricultural products for low cost and high timeliness of cold chain logistics. In terms of modeling, a multi-objective supply chain network optimization model with the minimum total cost, the minimum total carbon emission and the shortest distribution time is constructed, and carbon emission constraints are introduced into the model at the same time in order to respond to the requirements of the low-carbon strategy, and to take into account the operational efficiency of the company, the quality of the consumer service and the social and environmental benefits. In terms of algorithm design, a hybrid algorithm (LHS-SA-NSGA-II) is proposed, integrating Latin Hypercube Sampling (LHS), Simulated Annealing (SA), and the Non-Dominated Sorting Genetic Algorithm (NSGA-II). The algorithm improves population quality, accelerates convergence, and enhances local search by combining Latin Hypercube Sampling (LHS) with stochastic search for population initialization, thereby increasing diversity and uniformity in population distribution. Additionally, simulated annealing is introduced to optimize the offspring population after crossover and mutation operations, further enhancing population quality, accelerating convergence, and strengthening local search capabilities. The algorithm results show that, compared to the traditional NSGA-II and the SA-NSGA-II algorithms with random initialization, the LHS-SA-NSGA-II algorithm demonstrates clear advantages in terms of total cost, carbon emissions, and distribution time, confirming its superiority in optimizing cold chain networks. In addition, compared with the transfer mode, the constructed multi-modal coordinated distribution system significantly reduces the system operation cost and carbon emission level with a slight increase in the distribution time, which reflects its potential application in the decarbonization and cost control of the cold chain system. |
|---|---|
| ISSN: | 2199-4536 2198-6053 |