Detection of False Data Injection Attack in CPS by Adaptive Unscented Kalman Filter

Over the past few years, the use of Cyber-Physical Systems (CPS) has become more and more prevalent in various industries. With this increase in usage, there has been a surge of interest in exploring frequent network attacks and creating reliable techniques for detecting false data injection attacks (FDIA). Considering that the fake data injection attack can keep the residual unchanged before and after the attack, bypassing the traditional insufficient data monitoring technology. Based on the existing power grid model, the attack characteristics of FDIA are analyzed, and an improved algorithm is proposed based on the adaptive Kalman filter algorithm. The algorithm's effectiveness is tested by combining the existing Euclidean and Bayonet similarity detection methods.