Frequency Injection Based HVDC Attack-Defense Control Via Squeeze-Excitation Double CNN

Journal Article · Tue May 11 00:00:00 EDT 2021 · IEEE Transactions on Power Systems

DOI:https://doi.org/10.1109/tpwrs.2021.3078770· OSTI ID:1832697

Sun, Kaiqi ^[1]; Qiu, Wei ^[2]; Yao, Wenxuan ^[3]; You, Shutang ^[4]; Yin, He ^[4]; Liu, Yilu ^[5]

---

1. Shandong Univ. (China)
2. Hunan Univ., Changsha (China); Univ. of Tennessee, Knoxville, TN (United States)
3. Hunan Univ., Changsha (China)
4. Univ. of Tennessee, Knoxville, TN (United States)
5. Univ. of Tennessee, Knoxville, TN (United States); Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)

Due to the independent controllability and fast power regulation capability, the High Voltage Direct Current (HVDC) system could be a prospective technology to provide multiple ancillary services to the system besides conventional bulk power transmission. However, with the increase of False Data Injection Attacks (FDIAs) on PMU data, the HVDC system could have the wrong response once the collected data that the HVDC system relied on is attacked, thus threatening the system operating security. How to ensure the security of the PMU-based HVDC ancillary service control become an urgent issue. To mitigate the risk, this paper proposed an HVDC attack-defense control based on the FDIAs detection method. Firstly, the Squeeze-Excitation based Double Convolutional Neural Networks (SE-DCNN) is proposed to realize fast identification of the attacking frequency type based on the time and frequency domain signals. Furthermore, the duration time of FDIAs is detected by the local outlier factor. Then, utilizing the results from SE-DCNN, HVDC ancillary service control framework is reorganized and an HVDC attack defense control is proposed for suppressing the potential influence of various types of FDIAs on the HVDC system ancillary service. Different experiments results demonstrate that the proposed method has the ability to significantly mitigate the frequency deviation and oscillation under the FDIA.

• Research Organization: Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
• Sponsoring Organization: USDOE
• Grant/Contract Number: AC05-00OR22725
• OSTI ID: 1832697
• Journal Information: IEEE Transactions on Power Systems, Journal Name: IEEE Transactions on Power Systems Journal Issue: 6 Vol. 36; ISSN 0885-8950
• Publisher: IEEE
• Country of Publication: United States
• Language: English

Related Subjects

42 ENGINEERING
False Data Injection Attacks
HVDC ancillary service control
HVDC attack-defense control
Squeeze-Excitation based Double Convolutional Neural Networks