A Novel Data Analytical Approach for False Data Injection Cyber-Physical Attack Mitigation in Smart Grids

False data injection cyber-physical threat is a typical integrity attack in modern smart grids. These days, data analytical methods have been employed to mitigate false data injection attacks (FDIAs), especially when large scale smart grids generate huge amounts of data. In this paper, a novel data analytical method is proposed to detect FDIAs based on data-centric paradigm employing the margin setting algorithm (MSA). The performance of the proposed method is demonstrated through simulation using the six-bus power network in a wide area measurement system environment, as well as experimental data sets. Two FDIA scenarios, playback attack and time attack, are investigated. Experimental results are compared with the support vector machine (SVM) and artificial neural network (ANN). The results indicate that MSA yields better results in terms of detection accuracy than both the SVM and ANN when applied to FDIA detection.