Mitigating Attacks With Nonlinear Dynamics on Actuators in Cyber-Physical Mechatronic Systems

Abstract

The impact and mitigation of false data injection (FDI) attacks with nonlinear dynamics targeting actuators in cyber-physical mechatronic systems (CPMSs) is investigated in this paper. Actuators in mechatronic systems exhibit vulnerabilities to inputs with well-known nonlinearities (e.g., backlash, deadzone, and saturation), where the nonlinear dynamics can affect the actuators’ performance. A mitigation approach is proposed based on the retrospective cost-based adaptive control to stabilize and regulate the CPMS under such FDI cyberattack. Since mechatronic systems are implemented with actuators of different dynamical properties, this paper considers systems of linear and nonlinear dynamics. Simulation results demonstrate how the proposed adaptive control system achieves internal model control with the dynamics of the actuator systems and the nonlinearities of the backlash, deadzone, and saturation attacks. Results further show that the controller inverts and rejects the effects of attacks with unknown nonlinearities.