Resilient adaptive quantized control for nonlinear cyber-physical systems under deception attacks

For resilient adaptive control of nonlinear cyber-physical systems (NCPSs) subject to state-dependent malicious attacks, this article develops a dynamic-gain scaling method with novel coupled matrix inequalities, which is completely different from the popular various backstepping methods and has the following characteristics: (i) The stubborn assumption of differentiability of sensor attacks in the existing backstepping methods is removed, especially for dynamic output feedback control; (ii) Compared with the adaptive backstepping output feedback control results achieving semi-global uniform ultimate boundedness, this article proposes an adaptive quantized output feedback control method that can achieve global asymptotical state regulation for NCPSs under sensor attacks; (iii) Two dynamic gains are designed to compensate synergistically the uncertain nonlinearities and the unknown parameters under a non-identification mechanism, where the nonlinearities are allowed to be dependent of unmeasured states. In addition, the proposed control method of this article is applicable to the class of NCPSs with both actuator and sensor attacks, as well as any sector-bounded input quantizer.