Neuroadaptive Impulsive Control on Consensus of Uncertain Multiagent Systems Using Continuous and Sampled Information
Yiyan Han,Qiang Xiao,Zhigang Zeng
DOI: https://doi.org/10.1109/tnnls.2021.3126531
IF: 14.255
2023-01-01
IEEE Transactions on Neural Networks and Learning Systems
Abstract:This article considers the consensus problem of uncertain multiagent systems, which is addressed by neuroadaptive impulsive control schemes. The proposed control schemes indicate that the communication among agents only occurs impulsively, while the dynamics uncertainty is addressed by adaptive schemes using neural networks. Based on such approaches, two specific control schemes are designed. One is that with impulsive feedback, the control scheme uses continuous-time information, which implies that the adaptive process is continuous over time. Another is that by adopting sampled information, the update of all systems, including the feedbacks on agents, the update of neural networks, and the estimation for uncertainty, can be executed only at impulsive instants. The latter case can reduce the energy cost for communication and control, but extra assistant systems are required. The estimation and consensus prove to be achieved with errors if some conditions are fulfilled. Numerical simulations, including a practical system example, are presented.