Dynamic event-triggered adaptive neural control for MIMO nonlinear CPSs with time-varying parameters and deception attacks

This paper provides a dynamic event-triggered adaptive neural controller for the multi-input–multi-output (MIMO) non-strict feedback nonlinear cyber–physical systems (CPSs) with time-varying parameters encountering deception attacks (DAs). The single parameter learning method (SPLM) and the dynamic surface technology are combined under backstepping framework with adroitness, which not only resist the malicious sensor DA, but also reduce the complexity of the design scheme. Furthermore, a dynamic event-triggered strategy is added to update the threshold dynamically to further reduce the communication load. Theoretical analysis presents that the proposed control scheme can guarantee that all signals are bounded regardless encountering the DAs. Eventually, the validity of the developed methods is illustrated by a simulation case.