Frequency Control of Smart Grids Under Complex Hybrid Deception Attacks via Point-by-Point Model-Reference Tracking

Load Frequency Control (LFC) is critical for maintaining stability in smart grids (SGs) against complex hybrid (deterministic and/or stochastic) cyber-deception attacks. This paper presents a new approach to handle such a complicated problem via trajectory tracking of the response of a pre-designed ideal attack-free model. This method guarantees robust tracking in LFC and desired transient behavior (e.g., rising time, settling time). More precisely, the optimal attack-free model’s response is controlled by the chosen control strategy as a result of load variations. The output of such a model is used online to clean up the cyber-deception attacked output of the system. The cleaned output is then used to estimate the states of the system using the recently developed Regularized Least Squares (RLS) observer. Then, a control strategy is proposed to ensure a trajectory tracking mechanism where the cyber-attacked system is forced to follow the trajectories of an ideal attack-free reference model. Unlike other approaches, which attenuate the effect of cyber-attacks, the present approach eliminates the distortions created by these attacks completely. The stability of the proposed scheme is rigorously analyzed, and its efficacy is validated through the application of an isolated power system.