Weighted FDI Attack Detection With Uncertain Parameters in Nonlinear Cyber-Physical Systems

This research addresses the detection of weighted false data injection (FDI) attacks in nonlinear cyber-physical systems (CPSs), aiming to enhance both early warning capability and detection accuracy. First, considering the parameter uncertainties commonly present in real-world systems, two attack detection (AD) estimators are developed with varying degrees of conservatism, aiming to enhance detection accuracy and allow flexible selection according to different practical conditions. Second, a weighted method for detecting unknown FDI attack signals is proposed, significantly enhancing detection speed. Third, to address communication bandwidth limitations, a quantizer is employed to process transmission signals. This article aims to address the challenge of slow AD in CPSs, considering factors such as parameter uncertainties, packet loss, noise, and constraints on communication bandwidth. Finally, the effectiveness of the proposed method is validated through the longitudinal kinematics model of the aircraft.