Prior Knowledge-Augmented Broad Reinforcement Learning Framework for Fault Diagnosis of Heterogeneous Multiagent Systems

被引:6
作者
Guo, Li [1 ,2 ]
Ren, Yiran [1 ]
Li, Runze [3 ]
Jiang, Bin [3 ]
机构
[1] Anhui Polytech Univ, Sch Elect Engn, Wuhu 241000, Peoples R China
[2] Hubei Minzu Univ, Coll Intelligent Syst Sci & Engn, Enshi 445000, Peoples R China
[3] Nanjing Univ Aeronaut & Astronaut, Coll Automat Engn, Nanjing 211106, Peoples R China
基金
中国国家自然科学基金;
关键词
Feature extraction; Reinforcement learning; Fault diagnosis; Training; Multi-agent systems; Data mining; Bridges; Broad reinforcement learning (RL); distributed fault diagnosis (FD); heterogeneous multiagent systems (MASs); prior knowledge augmentation; INTERFERENCE;
D O I
10.1109/TCDS.2023.3266791
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
A heterogeneous multiagent system (MAS) can easily experience unpredictable faults due to its complex structure and involvement of different individuals. However, existing approaches have several issues, including complicated network architecture, insufficient feature extraction, and poor generalization ability. This study proposes a novel framework called prior knowledge-augmented broad reinforcement learning (PK-BRL) to effectively diagnose faults in a heterogeneous MAS. First, we construct a highly realistic visualized heterogeneous MAS and perform fault injection. Second, we present a novel fault diagnosis (FD) framework based on broad reinforcement learning (RL) with prior knowledge that effectively integrates offline RL and broad learning into the FD process. The interaction between heterogeneous multiagents and the constructed environment enables us to learn a superior FD strategy. Finally, experiments conducted on software-in-the-loop and hardware-in-the-loop platforms demonstrate that the proposed PK-BRL framework has a state-of-the-art diagnostic accuracy for the heterogeneous MAS, which offers valuable theoretical and practical significance for real-world application.
引用
收藏
页码:115 / 123
页数:9
相关论文
共 33 条
[1]   A multi-agent-based for fault location in distribution networks with wind power generator [J].
Azeroual, Mohamed ;
Boujoudar, Younes ;
Aljarbouh, Ayman ;
El Moussaoui, Hassan ;
El Markhi, Hassane .
WIND ENGINEERING, 2022, 46 (03) :700-711
[2]   Reinforcement learning multi-agent system for faults diagnosis of mircoservices in industrial settings [J].
Belhadi, Asma ;
Djenouri, Youcef ;
Srivastava, Gautam ;
Lin, Jerry Chun-Wei .
COMPUTER COMMUNICATIONS, 2021, 177 :213-219
[3]   Broad Learning System: An Effective and Efficient Incremental Learning System Without the Need for Deep Architecture [J].
Chen, C. L. Philip ;
Liu, Zhulin .
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2018, 29 (01) :10-24
[4]   Fixed-Time Fault-Tolerant Formation Control for a Cooperative Heterogeneous Multiagent System With Prescribed Performance [J].
Cheng, Wanglei ;
Zhang, Ke ;
Jiang, Bin .
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS, 2023, 53 (01) :462-474
[5]   Fixed-Time Fault-Tolerant Formation Control for Heterogeneous Multi-Agent Systems With Parameter Uncertainties and Disturbances [J].
Cheng, Wanglei ;
Zhang, Ke ;
Jiang, Bin ;
Ding, Steven X. .
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS, 2021, 68 (05) :2121-2133
[6]   RFlySim: Automatic test platform for UAV autopilot systems with FPGA-based hardware-in-the-loop simulations [J].
Dai, Xunhua ;
Ke, Chenxu ;
Quan, Quan ;
Cai, Kai-Yuan .
AEROSPACE SCIENCE AND TECHNOLOGY, 2021, 114
[7]   A Multi-Agent Approach Based on Machine-Learning for Fault Diagnosis [J].
El Koujok, Mohamed ;
Ragab, Ahmed ;
Amazouz, Mouloud .
IFAC PAPERSONLINE, 2019, 52 (10) :103-108
[8]   Domain Knowledge-Based Deep-Broad Learning Framework for Fault Diagnosis [J].
Feng, Jian ;
Yao, Yu ;
Lu, Senxiang ;
Liu, Yue .
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, 2021, 68 (04) :3454-3464
[9]   Formation Tracking of Multiagent Systems With Time-Varying Actuator Faults and Its Application to Task-Space Cooperative Tracking of Manipulators [J].
Feng, Zhi ;
Hu, Guoqiang .
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2023, 34 (03) :1156-1168
[10]   Fault Detection and Diagnosis Using Statistic Feature and Improved Broad Learning for Traction Systems in High-Speed Trains [J].
Guo L. ;
Li R. ;
Jiang B. .
IEEE Transactions on Artificial Intelligence, 2023, 4 (04) :679-688