人工神经网络
仿射变换
容错
非线性系统
控制理论(社会学)
计算机科学
多智能体系统
分布式计算
人工智能
控制(管理)
数学
物理
量子力学
纯数学
作者
Haiqing Wang,Jiuxiang Dong
出处
期刊:IEEE Transactions on Intelligent Transportation Systems
[Institute of Electrical and Electronics Engineers]
日期:2023-10-17
卷期号:: 1-9
被引量:2
标识
DOI:10.1109/tits.2023.3322689
摘要
Under a premise of universal rigidity, the affine formation control method based on stress matrix can solve the formation maneuvering problem well. However, a failure of the agent in the system can easily destroy this condition of rigidity. Consequently, a fault-tolerant affine formation control problem for multi-agent systems (MASs) with partial loss-of-effectiveness (PLOE) and bias faults is investigated. In this paper, a neural network-based hierarchical fault-tolerant affine formation (NN-HFAF) control strategy is proposed for heterogeneous nonlinear MASs. Firstly, some virtual systems are built as a link between leaders and followers. The virtual systems affinely locate their target positions in the formation maneuvers through the real-time positions of leaders. Then, an adaptive fault-tolerant control algorithm is designed for followers to tracking the virtual systems. It can effectively prevents the impact of a few agent failures from spreading to the whole network. And the system dynamics of agents in the network are considered to be heterogeneous. Moreover, radial basis function neural networks (RBF-NNs) are introduced to approximate the nonlinear functions of dynamic systems, the computational burden is reduced by adopting the single parameter learning mechanism. Finally, the numerical simulations are given to verify the efficiency of the proposed method.
科研通智能强力驱动
Strongly Powered by AbleSci AI