Observer-based data-driven attack compensation control algorithm for nonlinear systems with aperiodic DoS attacks

This article investigates the model-free adaptive tracking control problem of nonlinear systems under the aperiodic denial-of-service (DoS) attack. A new attack compensation framework composed of an observer and an output reconstruction mechanism is established. The observer and the output reconstruction mechanism work collaboratively and are updated based on current and historical data. With the proposed attack compensation framework, the assumptions about the boundness of attack compensation signals and the sign of pseudo-partial derivative (PPD) parameters can be removed. Further, an observer-based model-free adaptive compensation control (O-MFACC) algorithm with fewer conditions and assumptions is proposed to better achieve the tracking control objective and mitigate the impact of aperiodic DoS attacks. Finally, a simulation example is provided in comparison to validate the validity of the proposed method.