Author(s)

WeiBin Zhao, XinHai Xia^*

Affiliation(s)

School of Future Transportation, Guangzhou Maritime University Guangzhou, Guangzhou 510725, Guangdong, China.

Corresponding Author

XinHai Xia

ABSTRACT

Driven by rapid urbanization and motorization, peak-hour oversaturation has become a chronic issue in Guangzhou’s urban road network. Core districts such as Tianhe and Haizhu frequently experience congestion indices exceeding 7.8 during morning and evening peaks, with average vehicle speeds dropping below 19 km/h. Queue spillover triggers cascading “domino effects,” severely degrading system-wide traffic efficiency. While control subareas serve as fundamental units for regional signal coordination, conventional static partitioning fails to adapt to dynamic traffic fluctuations, and existing dynamic methods often neglect strategic interactions and interest conflicts among subareas, leading to suboptimal coordination.

This study proposes a game-theoretic framework for dynamic subarea adjustment under oversaturated conditions, using Guangzhou as a case study. We develop two decision-making paradigms: a centralized cooperative game model that minimizes total network control cost, and decentralized non-cooperative models—including Stackelberg and Nash equilibria—that account for autonomous subarea behavior. Through theoretical derivation, numerical simulation, and validation via the VISSIM microsimulation platform, we analyze equilibrium properties and system performance across different game structures.

Results demonstrate that the centralized model achieves the lowest total control cost—31.5% lower than the Nash equilibrium—but suffers from poor real-time responsiveness due to computational complexity. The Stackelberg model, leveraging a “central guidance-subarea response” mechanism, strikes an optimal trade-off: it reduces total delay by 18.2% compared to Nash, increases adjustment frequency by 112.5% relative to centralized control, and maintains robust performance under uncertainty. Meanwhile, the Nash model exhibits superior robustness (delay fluctuation ≤4.8% under ±20% data noise) but incurs significant efficiency losses.

Parametric analysis reveals that when the collaboration weight λ > 0.6, subarea merging probability increases by 42% and queue lengths decrease by 27%. Furthermore, constraining subarea size to 2-5 intersections optimally balances management overhead and coordination benefits—a finding validated across Guangzhou’s heterogeneous urban fabric. This research provides both theoretical grounding and a practical implementation pathway for intelligent, adaptive traffic control in oversaturated megacities, directly supporting Guangzhou’s “14th Five-Year Plan” for Intelligent Transportation Systems (ITS).

KEYWORDS

Oversaturated road network; Control subarea; Game equilibrium; Dynamic adjustment; Signal coordination

CITE THIS PAPER

WeiBin Zhao, XinHai Xia. Game-theoretic approach to dynamic control subarea adjustment in oversaturated urban road networks: evidence from Guangzhou. World Journal of Engineering Research. 2025, 3(5): 55-59. DOI: https://doi.org/10.61784/wjer3059.

REFERENCES

[1] Guangzhou Municipal Bureau of Transport. Annual Report on Transportation Development, 2024.

[2] Luo J, Zhang Q. Subdivision of Urban Traffic Area Based on the Combination of Static Zoning and Dynamic Zoning. Discrete Dynamics in Nature and Society, 2021(1): 9954267.

[3] Zhang S, Zhang R, Hou X, et al. A method for dynamic network partitioning based on intersection spatiotemporal similarity. Computer Simulation, 2024, 41(07): 189-194+221.

[4] Wang M, Yin C. Optimal Signal Timing Strategy for Road Intersection Based on Cooperative Game Theory. Proceedings of the 27th China Control and Decision Conference (Volume II), 2015: 2137-2142.

[5] Cheng P, Chu Y, Du Y. A Cooperative Model for Real-Time P2P Streaming under Game Theory Framework. Computer Applications, 2011, 31(05): 1159-1161+1188.

[6] Chen Y, Xu J. Vissim-based Traffic Simulation and Analysis of Guangzhou BRT System. Science, Technology and Engineering, 2010, 10(30): 7472-7476.

[7] Li S, Yoon HS. Sensor fusion-based vehicle detection and tracking using a single camera and radar at a traffic intersection. Sensors. 2023, 23(10): 4888.

[8] Yao W, Qian S. Learning to recommend signal plans under incidents with real-time traffic prediction. Transportation Research Record. 2020, 2674(6): 45-59.