Consensus tracking control for uncertain non‐strict feedback multi‐agent system under cyber attack via resilient neuroadaptive approach
Xiaoshan Ma,Lin Yang,Linru Ma,Wenhan Dong,Ming Jin,Long Zhang,Feng Yang,Yan Lin
DOI: https://doi.org/10.1002/rnc.6035
IF: 3.8973
2022-02-21
International Journal of Robust and Nonlinear Control
Abstract:Abstract Multi‐agent system (MAS) is a common cyber‐physical system (CPS). Due to it often uses a relatively open network platform, it is vulnerable to malicious cyber attacks in the process of system operation, so the control scheme design is critical to ensure the system operation security. In this article, we propose a new resilient neuroadaptive dynamic surface control scheme for non‐linear MASs with potential cyber attacks (false data injection and denial‐of‐service), system uncertainty, unknown control gain, and output constraints which are usually seen on CPSs. In this scheme, the neuroadaptive controller design ensures that all signals of the closed loop system are semi‐globally uniform ultimately bounded when the MAS even exists time‐varying cyber attacks on data links among agents, and the links between controllers, actuators, and sensors. Using Gaussian radial basis function neural network, the system uncertainty, non‐strict feedback terms, unknown control gain, and some cyber attacks are effectively solved and the process of controller design is extremely simplified. Finally, we provide a simulation result to verify the effectiveness and superiority of the proposed control scheme.