Adaptive Multi-Agent Reinforcement Learning with Graph Neural Networks for Dynamic Optimization in Sports Buildings
The dynamic scheduling optimization of sports facilities faces challenges posed by real-time demand fluctuations and complex interdependencies between facilities. To address the adaptability limitations of traditional centralized approaches, this study proposes a decentralized multi-agent reinforcem...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Buildings |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-5309/15/14/2554 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850076981167652864 |
|---|---|
| author | Sen Chen Xiaolong Chen Qian Bao Hongfeng Zhang Cora Un In Wong |
| author_facet | Sen Chen Xiaolong Chen Qian Bao Hongfeng Zhang Cora Un In Wong |
| author_sort | Sen Chen |
| collection | DOAJ |
| description | The dynamic scheduling optimization of sports facilities faces challenges posed by real-time demand fluctuations and complex interdependencies between facilities. To address the adaptability limitations of traditional centralized approaches, this study proposes a decentralized multi-agent reinforcement learning framework based on graph neural networks (GNNs). Experimental results demonstrate that in a simulated environment comprising 12 heterogeneous sports facilities, the proposed method achieves an operational efficiency of 0.89 ± 0.02, representing a 13% improvement over Centralized PPO, while user satisfaction reaches 0.85 ± 0.03, a 9% enhancement. When confronted with a sudden 30% surge in demand, the system recovers in just 90 steps, 33% faster than centralized methods. The GNN attention mechanism successfully captures critical dependencies between facilities, such as the connection weight of 0.32 ± 0.04 between swimming pools and locker rooms. Computational efficiency tests show that the system maintains real-time decision-making capability within 800 ms even when scaled to 50 facilities. These results verify that the method effectively balances decentralized decision-making with global coordination while maintaining low communication overhead (0.09 ± 0.01), offering a scalable and practical solution for resource management in complex built environments. |
| format | Article |
| id | doaj-art-fa5d3c2678264b6ebe5a60664cf9804c |
| institution | DOAJ |
| issn | 2075-5309 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Buildings |
| spelling | doaj-art-fa5d3c2678264b6ebe5a60664cf9804c2025-08-20T02:45:54ZengMDPI AGBuildings2075-53092025-07-011514255410.3390/buildings15142554Adaptive Multi-Agent Reinforcement Learning with Graph Neural Networks for Dynamic Optimization in Sports BuildingsSen Chen0Xiaolong Chen1Qian Bao2Hongfeng Zhang3Cora Un In Wong4School of Economics and Management, Xi’an Physical Education University, Xi’an 710068, ChinaFaculty of Humanities and Social Sciences, Macao Polytechnic University, Macao, ChinaCollege of Design, Hanyang University, Seoul 04763, Republic of KoreaFaculty of Humanities and Social Sciences, Macao Polytechnic University, Macao, ChinaFaculty of Humanities and Social Sciences, Macao Polytechnic University, Macao, ChinaThe dynamic scheduling optimization of sports facilities faces challenges posed by real-time demand fluctuations and complex interdependencies between facilities. To address the adaptability limitations of traditional centralized approaches, this study proposes a decentralized multi-agent reinforcement learning framework based on graph neural networks (GNNs). Experimental results demonstrate that in a simulated environment comprising 12 heterogeneous sports facilities, the proposed method achieves an operational efficiency of 0.89 ± 0.02, representing a 13% improvement over Centralized PPO, while user satisfaction reaches 0.85 ± 0.03, a 9% enhancement. When confronted with a sudden 30% surge in demand, the system recovers in just 90 steps, 33% faster than centralized methods. The GNN attention mechanism successfully captures critical dependencies between facilities, such as the connection weight of 0.32 ± 0.04 between swimming pools and locker rooms. Computational efficiency tests show that the system maintains real-time decision-making capability within 800 ms even when scaled to 50 facilities. These results verify that the method effectively balances decentralized decision-making with global coordination while maintaining low communication overhead (0.09 ± 0.01), offering a scalable and practical solution for resource management in complex built environments.https://www.mdpi.com/2075-5309/15/14/2554decentralized MARLgraph neural networksdynamic schedulingsports buildingsproximal policy optimizationuser satisfaction |
| spellingShingle | Sen Chen Xiaolong Chen Qian Bao Hongfeng Zhang Cora Un In Wong Adaptive Multi-Agent Reinforcement Learning with Graph Neural Networks for Dynamic Optimization in Sports Buildings Buildings decentralized MARL graph neural networks dynamic scheduling sports buildings proximal policy optimization user satisfaction |
| title | Adaptive Multi-Agent Reinforcement Learning with Graph Neural Networks for Dynamic Optimization in Sports Buildings |
| title_full | Adaptive Multi-Agent Reinforcement Learning with Graph Neural Networks for Dynamic Optimization in Sports Buildings |
| title_fullStr | Adaptive Multi-Agent Reinforcement Learning with Graph Neural Networks for Dynamic Optimization in Sports Buildings |
| title_full_unstemmed | Adaptive Multi-Agent Reinforcement Learning with Graph Neural Networks for Dynamic Optimization in Sports Buildings |
| title_short | Adaptive Multi-Agent Reinforcement Learning with Graph Neural Networks for Dynamic Optimization in Sports Buildings |
| title_sort | adaptive multi agent reinforcement learning with graph neural networks for dynamic optimization in sports buildings |
| topic | decentralized MARL graph neural networks dynamic scheduling sports buildings proximal policy optimization user satisfaction |
| url | https://www.mdpi.com/2075-5309/15/14/2554 |
| work_keys_str_mv | AT senchen adaptivemultiagentreinforcementlearningwithgraphneuralnetworksfordynamicoptimizationinsportsbuildings AT xiaolongchen adaptivemultiagentreinforcementlearningwithgraphneuralnetworksfordynamicoptimizationinsportsbuildings AT qianbao adaptivemultiagentreinforcementlearningwithgraphneuralnetworksfordynamicoptimizationinsportsbuildings AT hongfengzhang adaptivemultiagentreinforcementlearningwithgraphneuralnetworksfordynamicoptimizationinsportsbuildings AT corauninwong adaptivemultiagentreinforcementlearningwithgraphneuralnetworksfordynamicoptimizationinsportsbuildings |