Design of a Cross-border Trade Logistics Route Optimization System Based on Graph Neural Networks
Main Article Content
Keywords
cross-border trade logistics, route optimisation, graph neural networks, multi-objective decision-making, green logistics
Abstract
Against the backdrop of global digital trade development and the deepening of the Belt and Road Initiative, the complexity and dynamism of cross-border trade logistics networks continue to increase. Traditional route optimization methods are struggling to meet the core requirements of intelligent logistics, such as real-time response, multi-constraint coordination, and green and low-carbon development, thus becoming a bottleneck for the intelligent upgrading of cross-border logistics. Therefore, this paper designs a cross-border trade logistics path optimization system based on Graph Neural Network (GNN). Firstly, it refines the “departure–transit–customs clearance–destination” four-level structure of cross-border logistics, and constructs a logistics network graph model that integrates multiple dimensions of attributes. It then proposes the GCN-PPO hybrid optimisation algorithm, which uses GCN to extract network topology and spatio-temporal features, and combines this with PPO to enable multi-objective dynamic decision-making regarding cost, timeliness and environmental sustainability. Finally, a system with four core modules was developed based on Python and PyTorch. Numerical examples were designed based on simulated data to verify the effectiveness of the method. The results show that, compared with Dijkstra and NSGA-II algorithms, the system reduces transportation costs by 17.8% ± 1.2%, improves delivery timeliness by 23.5% ± 1.5%, reduces carbon emission intensity by 9.2% ± 0.6%, and achieves an order fulfillment rate of 96.8% ± 0.5%. This system has transcended the limitations of traditional methods and achieved intelligent and dynamic optimization of cross-border logistics routes. It provides a reusable technical solution for the intelligent, green and resilient upgrade of cross-border trade logistics.
References
- [1] Yang, H., Wu, P., Sang, M., et al. (2025). Research on logistics cost reduction strategies based on machine learning algorithms. Price: Theory & Practice, (3), 239–244.
- [2] Zhang, R., & Liu, X. (2021). Cloud logistics resource allocation based on spectral machine learning in big data environment. Statistics & Decision, 37(9), 177–179.
- [3] Qiao, Y. (2025). Empirical study on green transformation of logistics enterprises based on dual machine learning models. Logistics Technology, 44(6), 37–48.
- [4] Jiang, F. (2025). Research on optimization strategies of intelligent logistics systems based on machine learning. China Storage & Transport, (5), 80–81.
- [5] Fu, X., Bai, X., & Wei, B. (2022). Positioning accuracy detection algorithm for logistics automatic sorting robots based on machine learning. Automation and Instrumentation, (8), 245–248.
- [6] Li, C., Zhang, J., & Xu, W. (2012). Design and development of multi-site logistics industry business system. China Business and Market, 26(2), 39–43.
- [7] Ali, S. S., Khan, S., Fatma, N., Ozel, C., & Hussain, A. (2024). Utilisation of drones in achieving various applications in smart warehouse management. Benchmarking: An International Journal, 31(3), 920–954. https://doi.org/10.1108/BIJ-01-2023-0039.
- [8] Nawaf, Q. H. O., Yang, Q., & Jiang, X. (2026). Joint optimization of routing and resource allocation in decentralized UAV networks based on DDQN and GNN. Telecommunication Engineering, 66(1), 1–10.
- [9] Xiang, J., Song, S., Cao, W., et al. (2026). Physics-constrained graph neural networks for solving adjoint equations. Acta Mechanica Sinica, 42(1), 178–191.
- [10] Zhang, H., Wen, J., & Zhou, L. (2025). Reduced-order model of unsteady wind turbine wake based on a multifunctional recurrent fuzzy neural network. Journal of Southeast University (English Edition), 41(4), 437–445.
- [11] Zhan, B., Song, X., Cheng, J., et al. (2025). Pavement crack extraction based on multi-scale convolutional neural network. Transactions of Nanjing University of Aeronautics and Astronautics, 42(6), 749–766.
- [12] Churkin, A., Sauma, E., Pozo, D., et al. (2022). Enhancing the stability of coalitions in cross-border transmission expansion planning. IEEE Transactions on Power Systems, 37(4), 2744–2757.
- [13] Zhou, Q., & Bialek, J. W. (2005). Erratum: "Approximate model of European interconnected system as a benchmark system to study effects of cross-border trades" (IEEE Transactions on Power Systems). IEEE Transactions on Power Systems, 20(3), 1663.
- [14] Diallo, M. A. (2024). Design and implementation of data sovereignty and interoperability for data exchange in logistics through data space [Doctoral dissertation, Linyi University].
- [15] Bentum-Micah, G., Zhao, S., Ampofo, J., et al. (2025). E-business cooperation between China and West Africa: Investigating the risks and challenges. Acta Psychologica, 258, Article 105249.
- [16] Ma, D. V. F. M., Alcantara-Pilar, J. M., & Tolentino, M. (2025). Decoding the future: A bibliometric exploration of blockchain in logistics. Financial Innovation, 11(1), 1134–1157.
- [17] Wisnicki, B., Dzhuguryan, T., Mielniczuk, S., et al. (2025). An inventory management model for city multifloor manufacturing clusters under intermodal supply chain uncertainty. Sustainability, 17(21).
- [18] You, W., Lei, Y., Tan, H., et al. (2025). Conflict-free planning and data-driven control of large-scale nonlinear multi-robot systems. IET Cyber-Systems and Robotics, 7(3), 11–23.
- [19] Karaduman, O., & Gulhas, G. (2025). Blockchain-enabled supply chain management: A review of security, traceability, and data integrity amid the evolving systemic demand. Applied Sciences, 15(9), 5168.
Article Sidebar
This work is licensed under a Creative Commons Attribution 4.0 International License.