Adaptive control with trust-factor-guided strategy switching for resilient cyber-physical systems

Abstract Cyber-physical systems (CPS) are increasingly vulnerable to malicious sensor and actuator attacks due to their open networking architecture. This paper proposes an adaptive control framework augmented with a trust factor to improve resilience against such attacks. The trust factor provides a real-time measure of the system’s operating integrity and is used to adaptively weight multiple control strategies and to switch between them as conditions warrant. A nominal controller is combined with a robust/adaptive controller through an adaptive weighting determined by the trust factor. When the trust in the nominal control loop decreases (indicating a possible attack or fault), the framework shifts emphasis to the robust controller, and can seamlessly switch full control authority to it if necessary. All control laws, optimization criteria, and stability analysis are explicitly reformulated to include the trust factor. Lyapunov stability analysis is used to prove that the proposed trust-factor-integrated control scheme maintains boundedness of the closed-loop system and achieves attack mitigation. Simulation results on a representative nonlinear system (including an aircraft lateral–directional dynamics example) demonstrate that the trust-enhanced controller can successfully maintain stability and performance in the face of severe actuator attacks, outperforming a single-strategy approach.