Abstract
The integration of new technologies into the power grid leads to a growing, complex, interconnected system that is exposed to various cyber vulnerabilities. A power grid operating state can be altered due to the dynamic cyber-attacks which target different system objectives. This article brings forward the approach of power grid behavior analysis to identify two operating states: normal versus attacked. Once established the features for such states, we focus on Deep Neural Networks as security methods to mitigate the impact of cyber-attacks on the power grid by providing a case study simulation in MATLAB to sustain the proposed method.