Decentralized attack detection for multi‐area power systems via interconnection‐decoupled sliding mode observer

Kun Xu,Yugang Niu

DOI: https://doi.org/10.1002/rnc.6717

IF: 3.8973

2023-04-20

International Journal of Robust and Nonlinear Control

Abstract:This article deals with the false data injection attack (FDIA) detection problem for multi‐area power systems. Considering the large‐scale characteristic of the multi‐area power systems, a novel decentralized detection scheme is proposed to detect FDIAs. First, an unknown input sliding mode observer (UI‐SMO) is designed to decouple the interconnected terms, meanwhile, a robust term in the UI‐SMO can attenuate the effect of the load variation, by which the obtained residual is only sensitive to local FDIAs. The strict reachability of the sliding surface in the estimate error space is ensured by the robust term with an auxiliary variable, and the stability of the sliding mode dynamics is achieved. Then, a detection logic is delicately designed based on a two‐phase residual threshold. Due to the strong attenuating ability to load variation, the proposed detection scheme shows better detectability than some existing works with only decoupling interconnections. Finally, the theoretical results are backed up by a three‐area power system simulation with comparison between proposed UI‐SMO‐based and traditional UIO‐based detection scheme.

automation & control systems,engineering, electrical & electronic,mathematics, applied