From metasynthesis to edge intelligence: A novel entropy-regulated layered multi-agent coordination (ER-LMAC) paradigm for the management of open complex giant systems

Journals

Title：
From metasynthesis to edge intelligence: A novel entropy-regulated layered multi-agent coordination (ER-LMAC) paradigm for the management of open complex giant systems
Author：
Heng Ding1,2, Shimin Wu1,2,*, Guohui Du1,2, Rong Lu1,2 , Lijuan Su1 , Chuncheng Liu3,4, Kun Liu5 , Lijia Zhang3,4, Jinfeng Gao1,2
Author Affiliation：
¹Hebei Joy Smart Technology Co., Ltd., Shijiazhuang 050011, China
2Shenzhen Joy Smart Data Technology Co., Ltd., Shenzhen 518109, China
3Hebei Academy of Sciences, Shijiazhuang 050011, China
4Jingjinji National Center of Technology Innovation, Beijing 100094, China
5Hanjiang Zhixing Technology Co., Ltd., Xiangyang 441000, China
E-mail: *wushimin@outlook.com
Received：Nov. 24, 2025
Accepted：Dec. 5, 2025
Published：Dec.23, 2025

Abstract

Open complex giant systems (OCGSs) represent some of the most challenging problems in contemporary engineering and governance. Although Qian Xuesen’s OCGS methodology and metasynthesis system approach have provided a rich theoretical foundation, there remains a persistent gap between theory and scalable engineering practice, which we refer to as the entropy-increase dilemma. In this paper, we propose the Entropy-Regulated Layered Multi-Agent Coordination (ER-LMAC) paradigm as a modern, engineering-realizable framework for managing OCGSs. ER-LMAC synthesizes (i) entropy regulation as a cybernetic control objective, (ii) layered multi-agent coordination as a scalable decision mechanism, and (iii) edge intelligence implemented on an end–edge–cloud architecture as the computational substrate. As an empirical validation, we instantiate ER-LMAC in the domain of urban intelligent transportation systems by designing and deploying a decentralized, nonhistorical adaptive traffic signal control (ATSC) system in Xiangyang, China. The system covers 448 intersections in the main urban area and has operated continuously for more than three years. Longitudinal operational data indicate that road traffic efficiency increased by more than 20%, congestion duration decreased by more than 30%, and traffic accident rates were reduced by more than 60%. The city has maintained the lowest congestion index in its province for over 40 consecutive months, despite having one of the highest vehicle ownership levels. These results demonstrate that ER-LMAC can effectively resolve the entropy-increase dilemma in a real-world OCGS and suggest its broader applicability to domains such as smart grids, supply chain logistics, industrial Internet of Things (IIoT), and epidemic prevention and control.

Keywords

Open complex giant systems (OCGSs), entropy regulation, metasynthesis, layered multi-agent coordination, edge intelligence, adaptive traffic signal control, end–edge–cloud architecture, intelligent transportation systems (ITS).

References

[1] J. Gu, X. Tang, Some developments in the studies of meta-synthesis system approach, J. Syst. Sci. Syst. Eng. 12 (2025) 171–189. https://doi.org/10.1007/s11518-006-0128-4.

[2] Qian X., Additional Discussion on Open, Complex, Giant Systems, Pattern Recognit. Artif. Intell. 4 (1991) 1–4. https://sass.usst.edu.cn/2021/1102/c12541a258347/page.htm (accessed November 14, 2025).

[3] A. Stone, A Complex Systems Engineering Undertaking: The Qian Xuesen School of Systems Engineering, China Aerospace Studies Institute, 2024. https://www.airuniversity.af.edu/CASI/Display/Article/3643523/a-complex-systems-engineeringundertaking-the-qian-xuesen-school-of-systems-eng/.

[4] Longbing Cao, Ruwei Dai, Mengchu Zhou, Metasynthesis: M-space, M-interaction, and M-computing for open complex giant systems, IEEE Trans. Syst. Man Cybern. Part Syst. Hum. 39 (2009) 1007–1021. https://doi.org/10.1109/TSMCA.2009.2022414.

[5] X. Qian, J. Yu, R. Dai, A new discipline of science — the study of open complex giant system and its methodology, J. Syst. Eng. Electron. 4 (1993) 2–12. https://ieeexplore.ieee.org/document/6073209 (accessed May 6, 2025).

[6] Z. Gao, B. Jia, D. Xie, W. Wang, J. Wu, A Discussion on the Complexity and Transit Mechanisms of Urban Traffic Systems, Engineering 44 (2025) 24–29. https://doi.org/10.1016/j.eng.2024.12.006.

[7] R. Bayindir, E. Hossain, E. Kabalci, R. Perez, A comprehensive study on microgrid technology, Int. J. Renew. Energy Res. IJRER 4 (2014) 1094–1107. https://ijrer.ijrer.org/index.php/ijrer/article/view/2005 (accessed November 19, 2025).

[8] E.A. Hamza, B.E. Sedhom, E.A. Badran, Impact and assessment of the overvoltage mitigation methods in low‐voltage distribution networks with excessive penetration of PV systems: A review, Int. Trans. Electr. Energy Syst. 31 (2021) e13161. https://doi.org/10.1002/2050-7038.13161.

[9] M. Hanisch, C.M. Goldsby, N.E. Fabian, J. Oehmichen, Digital governance: A conceptual framework and research agenda, J. Bus. Res. 162 (2023) 113777. https://doi.org/10.1016/j.jbusres.2023.113777.

[10] O. Matthias, I. Fouweather, I. Gregory, A. Vernon, Making sense of big data – can it transform operations management?, Int. J. Oper. Prod. Manag. 37 (2017) 37–55. https://doi.org/10.1108/IJOPM-02-2015-0084.

[11] L. Cao, Data science: challenges and directions, Commun. ACM 60 (2017) 59–68. https://doi.org/10.1145/3015456.

[12] J. Gu, X. Tang, Meta-synthesis system approach to knowledge science, Int. J. Inf. Technol. Decis. Mak. 06 (2007) 559–572. https://doi.org/10.1142/S0219622007002629.

[13] Lu M., The Formation and Development of “Metasynthesis,” Strateg. Study CAE 7 (2005) 9–16. https://www.engineering.org.cn/sscae/EN/Y2005/V7/I1/9,.

[14] Dai R., Qian Xuesen’s View on Metasynthetic Engineering, Strateg. Study CAE 3 (2001) 14–20. https://www.engineering.org.cn/sscae/EN/Y2001/V3/I12/14.

[15] H. Tang, J. Liu, Z. Wang, Z. Gao, Projection exploration for multi-agent reinforcement learning, (2023). https://doi.org/10.21203/rs.3.rs-2759603/v1.

[16] Zheng N., Song Z., Ruwei D., Hall for Workshop of Metasynthetic Engineering Based on Man-Machine Integration, Pattern Recognit. Artif. Intell. 35 (2022) 767–773. http://manu46.magtech.com.cn/Jweb_prai/EN/Y2022/V35/I9/767.

[17] R. Zhang, G. Liu, Y. Liu, C. Zhao, J. Wang, Y. Xu, D. Niyato, J. Kang, Y. Li, S. Mao, S. Sun, X. Shen, D.I. Kim, Toward Edge General Intelligence with Agentic AI and Agentification: Concepts, Technologies, and Future Directions, (2025). https://doi.org/10.48550/arXiv.2508.18725.

[18] J. Zhou, F. Xiao, Y. Zhu, Hall for workshop of meta-synthetic engineering for complex product design, in: 2013 IEEE Int. Conf. Syst. Man Cybern., IEEE, Manchester, 2013: pp. 2173–2177. https://doi.org/10.1109/SMC.2013.372.

[19] Wang D., Zheng N., Liu C., Hall for workshop of metasynthetic engineering: the origin, development status and future, Acta Autom. Sin. 47 (2021) 1822–1839. https://doi.org/10.16383/j.aas.c210062.

[20] J. Wu, C.K.L. Or, Towards Open Complex Human-AI Agents Collaboration Systems for Problem Solving and Knowledge Management, (2025). https://doi.org/10.48550/arXiv.2505.00018.

[21] X. Wang, Q. Li, W. Jia, Cognitive Edge Computing: A Comprehensive Survey on Optimizing Large Models and AI Agents for Pervasive Deployment, (2025). https://doi.org/10.48550/arXiv.2501.03265.

[22] Y. Yu, Z. Zhai, W. Li, J. Ma, Target-oriented multi-agent coordination with hierarchical reinforcement learning, Appl. Sci. 14 (2024) 7084. https://doi.org/10.3390/app14167084.

[23] J. Mingers, C. Standing, What is information? Toward a theory of information as objective and veridical, J. Inf. Technol. 33 (2018) 85–104. https://doi.org/10.1057/s41265-017-0038-6.

[24] X. Jia, M. Guo, Y. Lyu, J. Qu, D. Li, F. Guo, Adaptive Traffic Signal Control Based on Graph Neural Networks and Dynamic Entropy-Constrained Soft Actor–Critic, Electronics 13 (2024) 4794. https://doi.org/10.3390/electronics13234794.

[25] Y. Yang, Y. Shi, X. Cui, J. Li, X. Zhao, A hybrid decision-making framework for UAV-assisted MEC systems: integrating a dynamic adaptive genetic optimization algorithm and soft actor–critic algorithm with hierarchical action decomposition and uncertainty-quantified critic ensemble, Drones 9 (2025) 206. https://doi.org/10.3390/drones9030206.

[26] X. Wei, Y. Xiang, J. Li, J. Liu, Wind power bidding coordinated with energy storage system operation in real-time electricity market: a maximum entropy deep reinforcement learning approach, Energy Rep. 8 (2022) 770–775. https://doi.org/10.1016/j.egyr.2021.11.216.

[27] D.R. Aleko, S. Djahel, An Efficient Adaptive Traffic Light Control System for Urban Road Traffic Congestion Reduction in Smart Cities, Information 11 (2020) 119. https://doi.org/10.3390/info11020119.

[28] S. Punitha, N.P. Subramaniam, P.A.D. Vimal Raj, A comprehensive review of microgrid challenges in architectures, mitigation approaches, and future directions, J. Electr. Syst. Inf. Technol. 11 (2024) 60. https://doi.org/10.1186/s43067-024-00188-4.

[29] N. Pöhner, M. Hennecke, Learning problem solving through educational robotics competitions: first results of an exploratory case study, in: Proc. 13th Workshop Prim. Second. Comput. Educ., ACM, Potsdam Germany, 2018: pp. 1–4. https://doi.org/10.1145/3265757.3265774.

[30] J. Dias, K. Althoefer, P.U. Lima, Robot competitions: what did we learn? [competitions], IEEE Robot. Autom. Mag. 23 (2016) 16–18. https://doi.org/10.1109/MRA.2015.2511678.

[31] S. Smith, G. Barlow, X.-F. Xie, Z. Rubinstein, Smart urban signal networks: initial application of the SURTRAC adaptive traffic signal control system, Proc. Int. Conf. Autom. Plan. Sched. 23 (2013) 434–442. https://doi.org/10.1609/icaps.v23i1.13594.

[32] C.H. Chen, M.F. Shiu, AgentFlow: Resilient Adaptive Cloud-Edge Framework for Multi-Agent Coordination, (2025). https://doi.org/10.36227/techrxiv.174681431.15296843/v1.

[33] H. Chen, W. Deng, S. Yang, J. Xu, Z. Jiang, E.C.H. Ngai, J. Liu, X. Liu, Toward edge general intelligence via large language models: opportunities and challenges, IEEE Netw. 39 (2025) 263–271. https://doi.org/10.1109/MNET.2025.3539338.

[34] M. Jiang, T. Wu, Z. Wang, Y. Gong, L. Zhang, R.P. Liu, A multi-intersection vehicular cooperative control based on end-edge cloud computing, IEEE Trans. Veh. Technol. 71 (2022) 2459–2471. https://doi.org/10.1109/TVT.2022.3143828.

[35] J. Liu, Y. Du, K. Yang, J. Wu, Y. Wang, X. Hu, Z. Wang, Y. Liu, P. Sun, A. Boukerche, V.C.M. Leung, Edge-Cloud Collaborative Computing on Distributed Intelligence and Model Optimization: A Survey, (2025). https://doi.org/10.48550/arXiv.2505.01821.

[36] H. Liang, L. Zhu, F.R. Yu, C. Yuen, Cloud-Edge-End Collaboration for Intelligent Train Regulation Optimization in TACS, IEEE Trans. Veh. Technol. 74 (2025) 454–465. https://doi.org/10.1109/TVT.2024.3462708.

[37] Y. Wang, J. Corey, Y. Lao, X. Xin, Criteria for the selection and application of advanced traffic signal control systems, 2012. https://rosap.ntl.bts.gov (accessed November 18, 2025).

[38] S. Mahato, Smart traffic signals: comparing MARL and fixed-time strategies, (2025). https://doi.org/10.48550/arXiv.2505.14544.

[39] N.H. Gartner, F.J. Pooran, C.M. Andrews, Implementation of the OPAC adaptive control strategy in a traffic signal network, in: ITSC 2001 2001 IEEE Intell. Transp. Syst. Proc. Cat No01th8585, 2001: pp. 195–200. https://doi.org/10.1109/ITSC.2001.948655.

[40] Xiangyang is china’s first city to have AI, IoV-powered traffic system, (n.d.). https://www.yicaiglobal.com/news/xiangyang-isfirst-chinese-city-to-introduce-ai-iov-powered-traffic-management-system (accessed November 19, 2025).

[41] J. Li, L. Peng, K. Hou, Y. Tian, Y. Ma, S. Xu, T.Z. Qiu, Adaptive signal control and coordination for urban traffic control in a connected vehicle environment: A review, Digit. Transp. Saf. 2 (2023) 89–111. https://doi.org/10.48130/DTS-2023-0008.

[42] Y. Wang, X. Yang, H. Liang, Y. Liu, A Review of the Self-Adaptive Traffic Signal Control System Based on Future Traffic Environment, J. Adv. Transp. 2018 (2018) 1–12. https://doi.org/10.1155/2018/1096123.

[43] K. Wu, J. Ding, J. Lin, G. Zheng, Y. Sun, J. Fang, T. Xu, Y. Zhu, B. Gu, Big-data empowered traffic signal control could reduce urban carbon emission, Nat. Commun. 16 (2025) 2013. https://doi.org/10.1038/s41467-025-56701-4.

[44] X. Huo, H. Huang, K.R. Davis, H.V. Poor, M. Liu, A review of scalable and privacy-preserving multi-agent frameworks for distributed energy resources, Adv. Appl. Energy 17 (2025) 100205. https://doi.org/10.1016/j.adapen.2024.100205.

[45] H. Zeng, Y. Huang, K. Zhan, Z. Yu, H. Zhu, F. Li, Multi-agent DDPG-based multi-device charging scheduling for IIoT smart grids, Sensors 25 (2025) 5226. https://doi.org/10.3390/s25175226.

[46] Q. Shen, L. Li, Acknowledged SoS architecture design based on agent-based modelling, in: D. Krob, L. Li, J. Yao, H. Zhang, X. Zhang (Eds.), Complex Syst. Des. Manag., Springer International Publishing, Cham, 2021: pp. 103–120. https://doi.org/10.1007/978-3-030-73539-5_9.

[47] H. Zhao, R. Xu, An Innovative Mechanism of Blockchain Technology on Joint Governance Model, J. Artif. Intell. Technol. 1 (2021) 84–91. https://doi.org/10.37965/jait.2020.0038.

[48] E. Pournaras, M.C. Ballandies, S. Bennati, C. Chen, Collective privacy recovery: data-sharing coordination via decentralized artificial intelligence, PNAS Nexus 3 (2024) 1–15. https://doi.org/10.1093/pnasnexus/pgae029.