Evaluation of false data injection attacks on automatic generation control considering wind power dynamics

Abstract

The automatic generation control (AGC) system is a crucial control system based on information and communication technology in interconnected power systems. However, the dependence on cyber systems has increased the risk of false data injection attacks (FDIAs) against AGC systems. Furthermore, as the penetration level of wind power increases, an AGC dynamic framework incorporating wind stochastic modeling should be developed, and the impact of FDIAs on AGC considering wind power penetrations should also be evaluated. This paper proposes an evaluation method of the impact of FDIAs on AGC considering wind power penetrations. First, a linearized analysis framework is developed, in which the network model is considered due to its geographical smoothing effects. Then, stochastic differential equations (SDEs) are employed to model wind power and load temporal uncertain behaviors, and the overall system dynamics under different FDIA types are also proposed. Finally, the proposed methodology is illustrated via a 4-area practical AGC system. The simulation results show the effects of different types of attacks on the AGC system considering wind power penetrations and demonstrate that the existing evaluation frameworks may overestimate FDIA effects and wind power fluctuation and corresponding system inertia reduction may help the attacker to cause system instability.