Data-Driven Containment Control for Unknown MIMO Nonlinear MASs Under Aperiodic DoS Attacks

This paper considers the problem of containment control for multi-input multi-output (MIMO) nonlinear multi-agent systems (MASs) with output saturation and aperiodic denial-of-service (DoS) attacks. By applying the dynamic linearization method, the nonlinear systems are modified to an equivalent linear data model. Based on the inputs and saturated outputs data, a new data-driven containment control algorithm in the presence of aperiodic DoS attacks is developed, which mainly includes the following perspectives: 1) based on the historical information of the MASs, a switched error system is designed to decrease the impact of aperiodic DoS attacks, where the time duration was constrained due to the limitation of attackers’ energy; 2) the relation between measured error and tracking error is constructed to deal with the difficulties of incomplete data suffered from output saturation; 3) with the help of contraction mapping method and mathematical induction approach, the boundedness of containment error is guaranteed by the convergence of the control strategy in such an insecure environment. Numerical simulations are given to validate the effectiveness of the framework. Note to Practitioners—This paper studies the containment control problem for MIMO nonlinear MASs, which has application value in robotic systems, intelligent transportation, multiple spacecraft systems, and other fields. Note that these systems are notably susceptible to network attacks and saturation nonlinearity, which may decrease the system performance or even cause instability. Since the agent dynamics generally are not accurate even unavailable, the existing model-based technologies are inappropriate for MIMO nonlinear MASs. Therefore, in light of the challenge aroused by aperiodic DoS attacks and output saturation, a novel switched error system-based model-free adaptive containment control approach is developed to preserve data integrity while the stability of containment error is guaranteed. Numerical simulations indicate the validity of the obtained results.