![[2602.17068] Spatio-temporal dual-stage hypergraph MARL for human-centric multimodal corridor traffic signal control](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.17068] Spatio-temporal dual-stage hypergraph MARL for human-centric multimodal corridor traffic signal control
Summary
This paper introduces STDSH-MARL, a novel framework for human-centric traffic signal control that enhances multimodal transportation efficiency using multi-agent reinforcement learning.
Why It Matters
As urban areas increasingly prioritize multimodal transportation, understanding and improving traffic signal control is crucial. This research provides a scalable solution that adapts to various traffic scenarios, emphasizing public transportation efficiency and user-centric design.
Key Takeaways
- STDSH-MARL utilizes a dual-stage hypergraph attention mechanism for improved traffic signal control.
- The framework enhances multimodal performance, particularly for public transportation.
- Experiments show superior performance compared to existing methods across various traffic scenarios.
- A hybrid discrete action space allows for adaptive signal timing decisions.
- Temporal hyperedges significantly contribute to performance improvements.
Computer Science > Machine Learning arXiv:2602.17068 (cs) [Submitted on 19 Feb 2026] Title:Spatio-temporal dual-stage hypergraph MARL for human-centric multimodal corridor traffic signal control Authors:Xiaocai Zhang, Neema Nassir, Milad Haghani View a PDF of the paper titled Spatio-temporal dual-stage hypergraph MARL for human-centric multimodal corridor traffic signal control, by Xiaocai Zhang and 2 other authors View PDF Abstract:Human-centric traffic signal control in corridor networks must increasingly account for multimodal travelers, particularly high-occupancy public transportation, rather than focusing solely on vehicle-centric performance. This paper proposes STDSH-MARL (Spatio-Temporal Dual-Stage Hypergraph based Multi-Agent Reinforcement Learning), a scalable multi-agent deep reinforcement learning framework that follows a centralized training and decentralized execution paradigm. The proposed method captures spatio-temporal dependencies through a novel dual-stage hypergraph attention mechanism that models interactions across both spatial and temporal hyperedges. In addition, a hybrid discrete action space is introduced to jointly determine the next signal phase configuration and its corresponding green duration, enabling more adaptive signal timing decisions. Experiments conducted on a corridor network under five traffic scenarios demonstrate that STDSH-MARL consistently improves multimodal performance and provides clear benefits for public transportation prio...
