Distributed Fuzzy Filtering for Load Frequency Control of Nonlinear Interconnected Power Systems under Cyber-Physical Attacks

In this study, a new distributed filtering approach is proposed for the load frequency control of uncertain non-linear power systems with cyber-physical attacks. Specifically, the non-linear power system is firstly modelled under interval type-2 Takagi–Sugeno fuzzy framework and the uncertainty therein is captured by designing corresponding membership functions. Both denial of service cyber attack and physical sensor attack are considered and modelled as independent Bernoulli process. Based on the Lyapunov stability theory, less conservative sufficient conditions have been derived to guarantee the robustly mean-square asymptotic stability with an average H ∞ performance standard γ for the dynamic filtering error system. Moreover, artful matrix transformation techniques have been adopted to decouple the intertwined matrix variables in designing the distributed filter gains. In simulations, a three-area non-linear power system with internal uncertainties is used to validate the robustness of the proposed distributed filtering strategy to system parametric uncertainties and cyber-physical attacks.