Neuro-adaptive consensus strategy for a class of nonlinear time-delay multi-agent systems with an unmeasurable high-dimensional leader

• Published in: IET control theory & applications Vol. 13; no. 2; pp. 230 - 238
• Main Authors: Yue, Dongdong, Cao, Jinde, Li, Qi, Shi, Xinli
• Format: Journal Article
• Language: English
• Published: The Institution of Engineering and Technology 29.01.2019
• Subjects: adaptive control; consensus approach; control system synthesis; delay systems; directed communication topology; distributed control; distributed cooperative consensus tracking problem; distributed robust radial basis function neural network controller; leader agent; local observer; Lyapunov methods; Lyapunov–Krasovskii functional method; matrix algebra; multi‐agent systems; multi‐robot systems; M‐matrix; neurocontrollers; neuro‐adaptive consensus strategy; nonlinear control systems; nonlinear time‐delay multiagent systems; observers; radial basis function networks; Research Article; robust control; time‐varying systems; tracking errors; uncertain multiagent systems; uncertain systems; unknown time‐varying delays; unmeasurable high‐dimensional leader; time-varying systems; M-matrix; multi-agent systems; delay systems; unmeasurable high-dimensional leader; multi-robot systems; uncertain multiagent systems; consensus approach; nonlinear time-delay multiagent systems; leader agent; neurocontrollers; radial basis function networks; robust control; Lyapunov–Krasovskii functional method; distributed control; distributed cooperative consensus tracking problem; uncertain systems; distributed robust radial basis function neural network controller; local observer; tracking errors; control system synthesis; adaptive control; unknown time-varying delays; matrix algebra; nonlinear control systems; directed communication topology; neuro-adaptive consensus strategy; observers; Lyapunov methods
• ISSN: 1751-8644; 1751-8652
• DOI: 10.1049/iet-cta.2018.5314

Distributed cooperative consensus tracking problem for a class of uncertain multi-agent systems with a high-dimensional leader under a directed communication topology is concerned. Compared with related works, the dynamics of the leader agent is allowed to be different from those of the followers and may not be measured. Meanwhile, the dynamics of each follower is subject to un-modelled dynamics, unknown time-varying delays, as well as external disturbances, which makes the model more suitable in various practical applications. Based on the $\textit {M}$M-matrix and Lyapunov–Krasovskii functional method, a distributed robust radial basis function neural network controller as well as a local observer is designed for each follower so as to guarantee the ultimate boundness of the tracking errors to the leader's output signals. By appropriately cutting down the neural network parameters to be updated online, the computational burden can be greatly reduced. The effectiveness of the proposed consensus approach is testified via a numerical example.