Model‐Free Resilient Tracking Control for MIMO Nonlinear CPSs Under Deception Attacks and Non‐Affine Faults

ABSTRACT The resilient tracking control of cyber‐physical systems (CPSs) against actuator faults and sensor attacks has been extensively researched. Most existing efforts have considered actuator faults and sensor attacks separately and primarily rely on models. Achieving effective tracking control under these circumstances poses a significant challenge for designers. To address this, a novel approach is presented that simultaneously addresses both actuator faults and sensor attacks. The model‐free methodology considers attack gains and utilizes compromised states to formulate appropriate controllers. Unlike previous studies, the approach considers nonaffine nonlinear faults. It is theoretically proven that the proposed tracking control scheme, employing a new coordinate transformation and a direct adaptive method without utilizing any Nussbaum technology, ensures that all signals in the closed‐loop system remain bounded and that the output can track the desired reference signal. Finally, simulation results confirm the feasibility of the proposed method.