Event-triggered neural adaptive anti-disturbance control of nonlinear multi-agent systems with asymmetric constraints

Abstract

The event-triggered adaptive control problem of stochastic nonlinear multi-agent systems subject to stochastic faults and asymmetric output constraints is investigated in this paper. Radial basis function neural networks (RBFNNs) are employed to adaptively approximate the unknown nonlinearities and changes in system dynamics model due to stochastic failures. By utilising a one-to-one nonlinear mapping, the asymmetric output constraints stochastic system is converted into a system without any constraints. Furthermore, to save the communication resources between controller and actuator, an improved dynamic event-triggered mechanism is developed, which contains threshold parameters and an exponential convergence term. Then, based on the stochastic Lyapunov function method, an event-triggered adaptive fault-tolerant controller is proposed for the considered systems. It is shown that the developed adaptive fault-tolerant controller can guarantee that all the signals remain semi-globally uniformly ultimately bounded while the output constraint is satisfied, even if the system is affected by stochastic failures. Eventually, the example results are provided to illustrate the effectiveness of the proposed control methodology.

Keywords:

• Event-triggered control
• nonlinear multi-agent systems
• neural networks (NNs)
• stochastic faults

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All data generated or analysed during this study are included in this published article.

Additional information

Funding

This work was partially supported by the National Natural Science Foundation of China [grant number 62073046].

Notes on contributors

Weidi Cheng

Weidi Cheng received the B.S. degree in automation from Nanning Normal University, Nanning, China, in 2018. He is currently pursuing the M.S. degree in control theory and control engineering with Bohai University, Jinzhou, China. His current research interests include adaptive fuzzy/neural control, event-triggered control, and multi-agent systems.

Hongjing Liang

Hongjing Liang received the B.S. degree in mathematics from Bohai University, Jinzhou, China, in 2009, the M.S. degree in fundamental mathematics from Northeastern University, Shenyang, China, in 2011, and ph.D. degree in control theory and control engineering from Northeastern University, Shenyang, China, in 2016. He was a Temporary Research Associate with the Science Program, Texas A\&M University, Doha, Qatar. He is currently an associate professor with the Bohai University, Jinzhou, China. His research interests include adaptive control, fuzzy control, multi-agent systems and their applications.

Shenglin Hu

Shenglin Hu received the B.S. degree in automation from Qingdao University of Technology, Qingdao, China, in 2019. He is currently pursuing the M.S. degree in control theory and control engineering with Bohai University, Jinzhou, China. His current research interests include multi-agent systems, adaptive finite-time control and event-triggered control.