Detecting bi-level false data injection attack based on time series analysis method in smart grid

Abstract

Smart grid is a crucial Cyber-Physical system and is prone to False Data Injection Attack (FDIA). In this paper, we propose a novel detection mechanism for a new-type FDIA which targets at inducing generation rescheduling and load shedding. We exploit a signal processing method to recognize the behavior features of the estimated states under this FDIA and employ the captured features to train a time-series-analysis based detector. Before training the detector, an improved ELM method is proposed to eliminate the redundancies of the feature vectors. By doing so, our proposed detection mechanism can effectively detect the new-type FDIA by analysing the deviations between the feature vectors in both the spatial and temporal aspects. We assess the performance of the proposed mechanism with comprehensive simulations on IEEE 14- and 118-bus systems. The results indicate that the proposed mechanism can be performed in a real-time way with satisfactory detection accuracy.