TY - GEN
T1 - SDHN
T2 - 2nd International Conference on Advanced Robotics, Control, and Artificial Intelligence, ICARCAI 2025
AU - Zhao, Delin
AU - Shan, Yanbo
AU - Liu, Chang
AU - Lin, Shenghang
AU - Shou, Yingxin
AU - Xu, Bin
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2027.
PY - 2027
Y1 - 2027
N2 - Multi-Agent Reinforcement Learning is widely used for multi-robot coordination, where simple graphs typically model pairwise interactions. However, such representations fail to capture higher-order collaborations, limiting effectiveness in complex tasks. While hypergraph-based approaches enhance cooperation, existing methods often generate arbitrary hypergraph structures and lack adaptability to environmental uncertainties. To address these challenges, we propose the Skewness-Driven Hypergraph Network (SDHN), which employs stochastic Bernoulli hyperedges to explicitly model higher-order multi-robot interactions. By introducing a skewness loss, SDHN promotes an efficient structure with Small-Hyperedge Dominant Hypergraph, allowing robots to prioritize localized synchronization while still adhering to the overall information, similar to human coordination. Extensive experiments on Moving Agents in Formation and Robotic Warehouse tasks validate SDHN’s effectiveness, demonstrating superior performance over state-of-the-art baselines.
AB - Multi-Agent Reinforcement Learning is widely used for multi-robot coordination, where simple graphs typically model pairwise interactions. However, such representations fail to capture higher-order collaborations, limiting effectiveness in complex tasks. While hypergraph-based approaches enhance cooperation, existing methods often generate arbitrary hypergraph structures and lack adaptability to environmental uncertainties. To address these challenges, we propose the Skewness-Driven Hypergraph Network (SDHN), which employs stochastic Bernoulli hyperedges to explicitly model higher-order multi-robot interactions. By introducing a skewness loss, SDHN promotes an efficient structure with Small-Hyperedge Dominant Hypergraph, allowing robots to prioritize localized synchronization while still adhering to the overall information, similar to human coordination. Extensive experiments on Moving Agents in Formation and Robotic Warehouse tasks validate SDHN’s effectiveness, demonstrating superior performance over state-of-the-art baselines.
KW - Hypergraph Neural Network
KW - Multi-Agent Reinforcement Learning
KW - Multi-Robot Coordination
KW - Skewness Loss
UR - https://www.scopus.com/pages/publications/105053596770
U2 - 10.1007/978-981-92-0971-2_13
DO - 10.1007/978-981-92-0971-2_13
M3 - 会议稿件
AN - SCOPUS:105053596770
SN - 9789819209705
T3 - Lecture Notes in Networks and Systems
SP - 135
EP - 147
BT - Proceedings of the Second International Conference on Advanced Robotics, Control, and Artificial Intelligence - ICARCAI 2025
A2 - Wang, Hai
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 24 November 2025 through 27 November 2025
ER -