Data-Driven Attack Detection and Resilient Control for Heterogeneous Nonlinear Nonaffine MASs

The data-driven attack detection and resilient control problems for heterogeneous nonlinear nonaffine multi-agent systems (MASs) in the presence of false data injection (FDI) attacks are investigated in this paper. The heterogeneous nonlinear nonaffine MASs are firstly translated to an equivalent linear data model by using the dynamic linearization method. Then, based on the obtained linear data model, a novel attack detector, by using the nonlinear autoregressive with exogenous input neural network (NARXNN), is designed to detect whether MASs are under FDI attacks. Further, a data-driven resilient switched control scheme with the attack compensation mechanism is designed to achieve the leaderless consensus control objective. Finally, a simulation is given to illustrate the efficiency of the proposed method by comparisons.