Resilient Output Regulation for Cyber-Physical Systems Against False Data Injection Attacks: A Hybrid Iteration Scheme.

Abstract

The linear optimal output regulation problem (LOORP) of discrete-time (DT) cyber-physical systems (CPSs) under false data injection attacks (FDIAs) is investigated in this article. First, the LOORP under FDIAs is reduced to a static optimization problem and a dynamic minimax problem, and the corresponding model-based schemes are provided to solve these two problems. Afterward, a hybrid iteration (HI)-based Q-learning scheme is proposed to solve the two issues online. This scheme requires neither exact system dynamics information nor an initially stabilizing control gain, which also achieves a fast iteration speed. Finally, a discretized $LCL$ -coupled inverter-based distributed generation system is presented to demonstrate the performance of the proposed scheme.