Mitigating the Impacts of False Data Injection Attacks in Smart Grids using Deep Convolutional Neural Networks

Abstract

The smart grid is vulnerable to cyberattacks due to the integration of information and communication technologies (ICT). The false data injection attack (FDIA) is a type of cyberattack that is against the state estimation of the power grid. It is imperative to mitigate the impacts of such a stealthy attack. In this paper, a deep convolutional neural network scheme was proposed. It has been evaluated on the IEEE 39-bus system using real-world load data and performs better than existed approaches.