An Adaptive LQR-Based Defense Strategy against False Data Injection Attack in Smart Grids

Abstract

With the rapid development of cyber-physical power system, the security risk caused by false data injection attacks on the power system is increasing. Due to the covert characteristics of false data injection, the existing detection methods can be deceived and lead to power system overload. To solve this problem, an adaptive defense method based on the linear-quadratic form (LQR) is proposed. Considering the impact of attack on the physical system, a physical dynamic model of the power grid is constructed. Taking the covert characteristics of false data injection attack into account, the adaptive LQR controller-based defense method is developed. Through the design of parameters of LQR, the proposed controller can respond quickly to deceptive attacks. Then, the developed LQR-based adaptive control method can ensure the stability of the system as soon as possible after being attacked. Finally, the performance of the proposed control method to restore the stability of power systems under false data injection attack is verified on the IEEE 5-bus.