False Data and Threat Detection Methods for Power Systems Under Cyber Attacks Based On Big Data and Deep Learning Algorithms

Author

Ali Parizad, Southern Illinois University CarbondaleFollow

Date of Award

9-1-2021

Degree Name

Doctor of Philosophy

Department

Electrical and Computer Engineering

First Advisor

Hatziadoniu, Constantine

Abstract

Power systems play a crucial role in the national economy and public safety of the countries. After the technological revolution in modern power systems, a large volume of data is incorporated into the processes of generation companies and customers. Consistent with this issue, we have the integration of Information and Communication Technologies (ICT) into the power system, making it a form of Cyber-Physical System (CPS). With the introduction of the ICT components, we have a new concept in the power system known as Big Data. Big data refers to a large volume of received data from gateways with distinct protocols. This interaction between information and power systems in different environments and specific protocols may lead to some loopholes between these two sections. In this case, various sectors such as information transmission, data collection, and control centers will be vulnerable and at risk of being attacks. These vulnerabilities originated from the cyber layer, resulting in security and stability issues in the power system. This dissertation addresses this challenging problem in modern power systems by employing Deep Neural Networks, Machine Learning Algorithms, and Information Theory concepts.