Journal
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
Volume 34, Issue 8, Pages 3979-3991Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TNNLS.2021.3121403
Keywords
Actuators; Fault tolerant systems; Fault tolerance; Control systems; Protocols; Artificial neural networks; Reinforcement learning; Containment control; fault-tolerant control (FTC); multiagent systems (MASs); reinforcement learning (RL)
Ask authors/readers for more resources
This article investigates the model-free fault-tolerant containment control problem for multiagent systems (MASs) with time-varying actuator faults. A distributed containment control method based on reinforcement learning (RL) is adopted to achieve the containment control objective without prior knowledge on the system dynamics. The article proposes an optimal regulation problem and employs the RL-based policy iteration method to deal with it, developing nominal and fault-tolerant controllers to compensate for actuator faults. Numerical simulations demonstrate the effectiveness and advantages of the proposed method.
This article investigates the model-free fault-tolerant containment control problem for multiagent systems (MASs) with time-varying actuator faults. Depending on the relative state information of neighbors, a distributed containment control method based on reinforcement learning (RL) is adopted to achieve containment control objective without prior knowledge on the system dynamics. First, based on the information of agent itself and its neighbors, a containment error system is established. Then, the optimal containment control problem is transformed into an optimal regulation problem for the containment error system. Furthermore, the RL-based policy iteration method is employed to deal with the corresponding optimal regulation problem, and the nominal controller is proposed for the original fault-free system. Based on the nominal controller, a fault-tolerant controller is further developed to compensate for the influence of actuator faults on MAS. Meanwhile, the uniform boundedness of the containment errors can be guaranteed by using the presented control scheme. Finally, numerical simulations are given to show the effectiveness and advantages of the proposed method.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available