False data injection attacks (FDIAs) pose persistent threats to the security of modern power systems by compromising data integrity through the manipulation of measurements. While data-driven detection models can effectively identify these attacks under stable grid topologies, ensuring accurate results, their performance significantly deteriorates when the grid topology changes due to events like fault restoration, routine maintenance, or power flow redistribution. This lack of adaptability in traditional data-driven methods leads to a substantial decline in detection accuracy in such dynamic environments. Addressing this challenge, this study proposes a dynamic topology change-aware FDIA detection method based on graph attention networks, named the AST-TGT model. The model incorporates two key sub-modules: an attention mechanism module and a temporal convolutional module, to address the spatial interference effects arising from dynamic topological changes. First, a spatio-temporal attention mechanism was introduced to enhance the representational capacity of network nodes by dynamically assigning attention weights, thereby improving the model's understanding of both the topological structure and measurement data. Subsequently, a spatio-temporal convolutional block, composed of temporal convolutional layers and graph attention layers, was proposed to capture the joint temporal and spatial dependencies of the power grid topology and dynamic operational data. Extensive experiments conducted on the IEEE 14-bus and IEEE 118-bus standard test systems demonstrate that, compared to other data-driven detection models, the proposed model effectively improves the accuracy and stability of FDIA detection in the face of grid topology changes and exhibits high robustness in complex interference environments.
Citation: Xing Liu, Weicheng Shen, Fengyong Li. A graph attention network-based framework for dynamic topology change-aware FDIA detection in smart grids[J]. AIMS Energy, 2025, 13(5): 1219-1240. doi: 10.3934/energy.2025045
False data injection attacks (FDIAs) pose persistent threats to the security of modern power systems by compromising data integrity through the manipulation of measurements. While data-driven detection models can effectively identify these attacks under stable grid topologies, ensuring accurate results, their performance significantly deteriorates when the grid topology changes due to events like fault restoration, routine maintenance, or power flow redistribution. This lack of adaptability in traditional data-driven methods leads to a substantial decline in detection accuracy in such dynamic environments. Addressing this challenge, this study proposes a dynamic topology change-aware FDIA detection method based on graph attention networks, named the AST-TGT model. The model incorporates two key sub-modules: an attention mechanism module and a temporal convolutional module, to address the spatial interference effects arising from dynamic topological changes. First, a spatio-temporal attention mechanism was introduced to enhance the representational capacity of network nodes by dynamically assigning attention weights, thereby improving the model's understanding of both the topological structure and measurement data. Subsequently, a spatio-temporal convolutional block, composed of temporal convolutional layers and graph attention layers, was proposed to capture the joint temporal and spatial dependencies of the power grid topology and dynamic operational data. Extensive experiments conducted on the IEEE 14-bus and IEEE 118-bus standard test systems demonstrate that, compared to other data-driven detection models, the proposed model effectively improves the accuracy and stability of FDIA detection in the face of grid topology changes and exhibits high robustness in complex interference environments.
| [1] |
Li F, Shen W, Bi Z, et al. (2024) Sparse Adversarial Learning for FDIA Attack Sample Generation in Distributed Smart Grids. Comput Model Eng Sci (CMES) 139: 2. https://doi.org/10.32604/cmes.2023.044431 doi: 10.32604/cmes.2023.044431
|
| [2] |
Li K, Li F, Wang B, et al. (2024) False data injection attack sample generation using an adversarial attention-diffusion model in smart grids. AIMS Energy 12: 1271–1293. https://doi.org/10.3934/energy.2024058 doi: 10.3934/energy.2024058
|
| [3] |
Han Y, Feng H, Li K, et al. (2023) False data injection attacks detection with modified temporal multi-graph convolutional network in smart grids. Comput Secur 124: 103016. https://doi.org/10.1016/j.cose.2022.103016 doi: 10.1016/j.cose.2022.103016
|
| [4] |
Liang G, Zhao J, Luo F, et al. (2016) A review of false data injection attacks against modern power systems. IEEE Trans Smart Grid 8: 1630–1638. https://doi.org/10.1109/TSG.2015.2495133 doi: 10.1109/TSG.2015.2495133
|
| [5] | Liu L, Esmalifalak M, Ding Q, et al. (2014) Detecting false data injection attacks on power grid by sparse optimization. IEEE Trans Smart Grid 5: 612–621. |
| [6] |
Rashed M, Kamruzzaman J, Gondal I, et al. (2022) False data detection in a clustered smart grid using unscented Kalman filter. IEEE Access 10: 78548–78556. https://doi.org/10.1109/ACCESS.2022.3193781 doi: 10.1109/ACCESS.2022.3193781
|
| [7] |
Chen Y, Huang S, Liu F, et al. (2018) Evaluation of reinforcement learning-based false data injection attack to automatic voltage control. IEEE Trans Smart Grid 10: 2158–2169. https://doi.org/10.1109/TSG.2018.2790704 doi: 10.1109/TSG.2018.2790704
|
| [8] |
Živković N, Sarić AT (2018) Detection of false data injection attacks using unscented Kalman filter. J Mod Power Syst Clean Energy 6: 847–859. https://doi.org/10.1007/s40565-018-0413-5 doi: 10.1007/s40565-018-0413-5
|
| [9] |
Singh SK, Khanna K, Bose R, et al. (2017) Joint-transformation-based detection of false data injection attacks in smart grid. IEEE Trans Ind Inf 14: 89–97. https://doi.org/10.1109/TII.2017.2720726 doi: 10.1109/TII.2017.2720726
|
| [10] |
Moslemi R, Mesbahi A, Velni JM (2017) A fast, decentralized covariance selection-based approach to detect cyber attacks in smart grids. IEEE Trans Smart Grid 9: 4930–4941. https://doi.org/10.1109/TSG.2017.2675960 doi: 10.1109/TSG.2017.2675960
|
| [11] |
Chakhchoukh Y, Lei H, Johnson BK (2019) Diagnosis of outliers and cyber attacks in dynamic PMU-based power state estimation. IEEE Trans Power Syst 35: 1188–1197. https://doi.org/10.1109/TPWRS.2019.2939192 doi: 10.1109/TPWRS.2019.2939192
|
| [12] |
Abbaspour A, Sargolzaei A, Forouzannezhad P, et al. (2019) Resilient control design for load frequency control system under false data injection attacks. IEEE Trans Ind Electron 67: 7951–7962. https://doi.org/10.1109/TIE.2019.2944091 doi: 10.1109/TIE.2019.2944091
|
| [13] |
Kurt MN, Yılmaz Y, Wang X (2019) Secure distributed dynamic state estimation in wide-area smart grids. IEEE Trans Inf Forensics Secur 15: 800–815. https://doi.org/10.1109/TIFS.2019.2928207 doi: 10.1109/TIFS.2019.2928207
|
| [14] |
Wang H, Ruan J, Ma Z, et al. (2019) Deep learning aided interval state prediction for improving cyber security in energy internet. Energy 174: 1292–1304. https://doi.org/10.1016/j.energy.2019.03.009 doi: 10.1016/j.energy.2019.03.009
|
| [15] |
Yang T, Murguia C, Kuijper M, et al. (2020) An unknown input multiobserver approach for estimation and control under adversarial attacks. IEEE Trans Control Network Syst 8: 475–486. https://doi.org/10.1109/TCNS.2020.3029160 doi: 10.1109/TCNS.2020.3029160
|
| [16] |
Ding Y, Ma K, Pu T, et al. (2021) A deep learning-based classification scheme for cyber-attack detection in power system. IET Energy Syst Integr 3: 274–284. https://doi.org/10.1049/esi2.12034 doi: 10.1049/esi2.12034
|
| [17] |
Ding Y, Ma K, Pu T, et al. (2021) A deep learning-based classification scheme for false data injection attack detection in power system. Electronics 10: 1459. https://doi.org/10.3390/electronics10121459 doi: 10.3390/electronics10121459
|
| [18] |
Wang Y, Zhang Z, Ma J, et al. (2021) KFRNN: An effective false data injection attack detection in smart grid based on Kalman filter and recurrent neural network. IEEE Int Things J 9: 6893–6904. https://doi.org/10.1109/JIOT.2021.3113900 doi: 10.1109/JIOT.2021.3113900
|
| [19] |
Lin WT, Chen G, Huang Y (2022) Incentive edge-based federated learning for false data injection attack detection on power grid state estimation: A novel mechanism design approach. Appl Energy 314: 118828. https://doi.org/10.1016/j.apenergy.2022.118828 doi: 10.1016/j.apenergy.2022.118828
|
| [20] |
Shabbir A, Manzoor HU, Zoha A, et al. (2025) Smart grid security through fusion-enhanced federated learning against adversarial attacks. Eng Appl Artif Intell 157: 111169. https://doi.org/10.1016/j.engappai.2025.111169 doi: 10.1016/j.engappai.2025.111169
|
| [21] | Zhang W, Deng C, Su X, et al. (2022) Spatial-temporal attention based interpretable deep framework for FDIA detection in smart grid. 2022 IEEE Sustainable Power and Energy Conference (iSPEC) https://doi.org/10.1109/iSPEC54162.2022.10032978 |
| [22] |
Elahi E, Anwar S, Al-kfairy M, et al. (2025) Graph attention-based neural collaborative filtering for item-specific recommendation system using knowledge graph. Expert Syst Appl 266: 126133. https://doi.org/10.1016/j.eswa.2024.126133 doi: 10.1016/j.eswa.2024.126133
|
| [23] |
Boyaci O, Narimani MR, Davis KR, et al. (2021) Joint detection and localization of stealth false data injection attacks in smart grids using graph neural networks. IEEE Trans Smart Grid 13: 807–819. https://doi.org/10.1109/TSG.2021.3117977 doi: 10.1109/TSG.2021.3117977
|
| [24] |
Li X, Wang Y, Lu Z (2023) Graph-based detection for false data injection attacks in power grid. Energy 263: 125865. https://doi.org/10.1016/j.energy.2022.125865 doi: 10.1016/j.energy.2022.125865
|
| [25] |
Su X, Deng C, Yang J, et al. (2024) DAMGAT-Based interpretable detection of false data injection attacks in smart grids. IEEE Trans Smart Grid 15: 4182–4195. https://doi.org/10.1109/TSG.2024.3364665 doi: 10.1109/TSG.2024.3364665
|
| [26] | Huang S, Li F, Li K, et al. (2025) Dependency-Aware GraphSAGE based interpretable FDIA detection using BiLSTM with SE-Attention in smart grids. IEEE Int Things J https://doi.org/10.1109/JIOT.2025.3578369 |
| [27] |
Li X, Hu L, Lu Z (2024) Detection of false data injection attack in power grid based on spatial-temporal transformer network. Expert Syst Appl 238: 121706. https://doi.org/10.1016/j.eswa.2023.121706 doi: 10.1016/j.eswa.2023.121706
|
| [28] |
Qu Z, Dong Y, Li Y, et al. (2024) Localization of dummy data injection attacks in power systems considering incomplete topological information: A spatio-temporal graph wavelet convolutional neural network approach. Appl Energy 360: 122736. https://doi.org/10.1016/j.apenergy.2024.122736 doi: 10.1016/j.apenergy.2024.122736
|
| [29] |
Wang J, Si Y, Zhu Y, et al. (2024) Cyberattack detection for electricity theft in smart grids via stacking ensemble GRU optimization algorithm using federated learning framework. Int J Electr Power Energy Syst 157: 109848. https://doi.org/10.1016/j.ijepes.2024.109848 doi: 10.1016/j.ijepes.2024.109848
|
| [30] |
Liao W, Zhu R, Yang Z, et al. (2023) Electricity theft detection using dynamic graph construction and graph attention network. IEEE Trans Ind Inf 20: 5074–5086. https://doi.org/10.1109/TII.2023.3331131 doi: 10.1109/TII.2023.3331131
|
| [31] |
Wu Y, Zu T, Guo N, et al. (2023) Laplace-domain hybrid distribution model based FDIA attack sample generation in smart grids. Symmetry 15: 1669. https://doi.org/10.3390/sym15091669 doi: 10.3390/sym15091669
|
| [32] |
Zheng Q, Zheng J, Mei F, et al. (2023) TCN-GAT multivariate load forecasting model based on SHAP value selection strategy in integrated energy system. Front Energy Res 11: 1208502. https://doi.org/10.3389/fenrg.2023.1208502 doi: 10.3389/fenrg.2023.1208502
|
Figures(4) / Tables(6)
Xing Liu, Weicheng Shen, Fengyong Li. A graph attention network-based framework for dynamic topology change-aware FDIA detection in smart grids[J]. AIMS Energy, 2025, 13(5): 1219-1240. doi: 10.3934/energy.2025045
DownLoad: