To optimize resource allocation in free trade zones (FTZs), this study proposes a computational synergy mechanism leveraging multimodal network modeling. We integrate five-dimensional data streams—capital flows, logistics networks, human mobility, institutional policies, and digital information—through multimodal feature fusion. Key innovations include: (1) A dynamic graph embedding framework using graph attention networks (GATs) to model cross-factor dependencies; (2) A multi-agent coordination algorithm combining Nash bargaining with proximal policy optimization (PPO) for resource conflict resolution; (3) Real-time resource scheduling via heuristic search and stream processing. Validated with China’s pilot FTZ operational data, cross-border clearance workflows accelerated by 32.4% through dynamic scheduling.
ChanghongP. The tenth anniversary of free trade pilot zones in China: fundamental experiences and enhancement strategies. China Economist2024; 19: 2–20.
2.
HuangFZhangXXuJ, et al.Network embedding by fusing multimodal contents and links. Knowl Base Syst2019; 171: 44–55.
3.
AlsrehinNOKlaibAFMagablehA. Intelligent transportation and control systems using data mining and machine learning techniques: a comprehensive study. IEEE Access2019; 7: 49830–49857.
4.
MaXHeZYangP, et al.Agent-based modelling and simulation for life-cycle airport flight planning and scheduling. J Simulat2024; 18: 15–28.
5.
GaoJLiPChenZ, et al.A survey on deep learning for multimodal data fusion. Neural Comput2020; 32: 829–864.
6.
WangYHuangWSunF, et al.Deep multimodal fusion by channel exchanging. Adv Neural Inf Process Syst2020; 33: 4835–4845.
7.
ZhaoHMiJLiangM. A multi-granularity information fusion method based on logistic regression model and Dempster-Shafer evidence theory and its application. Int J Mach Learn Cybern2022; 13: 3131–3142.
8.
ZhuHWangZShiY, et al.Multimodal fusion method based on self-attention mechanism. Wireless Commun Mobile Comput2020; 2020: 8843186–8843188.
9.
MunirABlaschEKwonJ, et al.Artificial intelligence and data fusion at the edge. IEEE Aero Electron Syst Mag2021; 36: 62–78.
10.
FengQLiuZChenCP. Broad and deep neural network for high-dimensional data representation learning. Inf Sci2022; 599: 127–146.
11.
YuYSiXHuC, et al.A review of recurrent neural networks: LSTM cells and network architectures. Neural Comput2019; 31: 1235–1270.
12.
ZhengZHuangSWengR, et al.Improving self-attention networks with sequential relations. IEEE/ACM Trans Audio Speech Lang Process2020; 28: 1707–1716.
13.
KhanSNaseerMHayatM, et al.Transformers in vision: a survey. ACM Comput Surv2022; 54: 1–41.
14.
ChenSZhangYYangQ. Multi-task learning in natural language processing: an overview. ACM Comput Surv2024; 56: 1–32.
15.
ZawishMDavySAbrahamL. Complexity-driven model compression for resource-constrained deep learning on edge. IEEE Trans Artif Intell2024; 5: 3886–3901.
16.
AdhikariP. Multi-modal clinical document understanding via joint text–image representations. J Big Data Process, Stream Analytics, and Real-Time Insights2024; 14: 16–29.
17.
PashamSD. Graph-based algorithms for optimizing data flow in distributed cloud architectures. International Journal of Acta Informatica2022; 1: 67–95.
18.
SamMLSaptariASallehMRB, et al.Comparison between linear programming and integer linear programming: a review. Int J Math Oper Res2018; 13: 91–106.
19.
TheveninSAdulyasakYCordeauJF. Stochastic dual dynamic programming for multiechelon lot sizing with component substitution. Inf J Comput2022; 34: 3151–3169.
20.
DavisMJLuYSharmaM, et al.Stochastic optimization models for workforce planning, operations, and risk management. Serv Sci2018; 10: 40–57.
21.
Vinod ChandraSSAnandHS. Nature inspired meta heuristic algorithms for optimization problems. Computing2022; 104: 251–269.
22.
MauludDHAmeenSYOmarN, et al.Review on natural language processing based on different techniques. AJRCoS2021; 10: 1–17.
23.
BaltrušaitisTAhujaCMorencyLP. Multimodal machine learning: a survey and taxonomy. IEEE Trans Pattern Anal Mach Intell2018; 41: 423–443.
24.
TayeMM. Framework for enhanced ontology alignment using BERT-based. IJRITCC2023; 11: 2853–2861.
25.
RaoTLiXXuM. Learning multi-level deep representations for image emotion classification. Neural Process Lett2020; 51: 2043–2061.
26.
NathDAnkitNDR, et al.Application of machine learning and deep learning in finite element analysis: a comprehensive review. Arch Comput Methods Eng2024; 31: 2945–2984.
27.
EktefaieYDasoulasGNooriA, et al.Multimodal learning with graphs. Nat Mach Intell2023; 5: 340–350.
28.
ChoiJHKimJEKimKT. Deep learning approach for radar-based people counting. IEEE Internet Things J2021; 9: 7715–7730.
29.
HeJLiH. A review of trade geography research: theoretical foundations, spatial patterns, and frontier dynamics. Geographical Research Bulletin2025; 4: 96–119.
30.
LeeMYuGLiGY. Graph embedding-based wireless link scheduling with few training samples. IEEE Trans Wireless Commun2020; 20: 2282–2294.
31.
SaputriTRDLeeSW. The application of machine learning in self-adaptive systems: a systematic literature review. IEEE Access2020; 8: 205948–205967.
32.
EskandarianAWuCSunC. Research advances and challenges of autonomous and connected ground vehicles. IEEE Trans Intell Transport Syst2019; 22: 683–711.
33.
TantalakiNSouravlasSRoumeliotisM. A review on big data real-time stream processing and its scheduling techniques. Int J Parallel, Emergent Distributed Syst2020; 35: 571–601.
34.
El-SabekLMMcCabeBY. Coordination challenges of production planning in the construction of international mega-projects in the Middle East. Int J Construct Educ Res2018; 14: 118–140.
