2nd International Graduate Research Symposium – IGRS’23, İstanbul, Türkiye, 2 - 05 Mayıs 2023, cilt.1, ss.342-351
State-estimation is a critical process for ensuring the secure and reliable operation of the power
systems by determination of the operating state of the system based on available measurements.
Recent researches show that state estimation process can be misled by False Data Injection
Attacks (FDIA). To conduct FDIA, attack vectors are injected to the compromised
measurements to bypass the bad data detection methods. While traditional state estimation is
already surmountable by attackers, with high penetration of distributed energy resources
(DERs), the process has become more vulnerable to the cyber-attacks. In this study, to detect
the cyber-attacks in power systems with high penetration rate of DERs, a Deep Learning Based
FDIA detection method is proposed which aims to detect the anomalies on state estimation and
measurements and identify the targeted bus by an attacker. The proposed approach is evaluated
on the IEEE 14 bus system using load data of the New York independent system operator from
January 2012 to October 2012. The proposed method is tested on four different system
configurations; No DERs and different DERs levels. Tested FDIA attacks conducted to the
system, target the system variables and manipulate system state variables. Attacked and nonattacked measurements and estimated state variables by state estimation are used to train the
proposed algorithm. Results indicates the reliability of the proposed method against detecting
the FDIAs on power systems with highly penetrated rate of DERs.