Mitigation of adversarial attacks: monitoring smart grids

These codes presents a deep learning approach based robust data engineering for mitigation of adversarial attacks and wide area monitoring.

These files describe an experiment performed on phasor measurement unites dataset that is made publicly available . The goal of the experiment is to train a deep network to be resilient against any adversarial attacks. A specific Robust feature engineering and a deep learning are involved in model reconstructions. fast gradient sign method and basic iterative method are involved in this case.

Notes: (i) To be able to produce experiments provided in of these codes, you have to run the "*.m" files in the directory in alphabetical order. (ii) Then you can plot results starting by any "plot...*.m" files.

link to the original paper:

https://www.researchgate.net/publication/371314420_Towards_Resilient_and_Secure_Smart_Grids_against_PMU_Adversarial_Attacks_A_Deep_Learning-Based_Robust_Data_Engineering_Approach

Please cite our work as:

Berghout, T.; Benbouzid, M.; Amirat, Y. Towards Resilient and Secure Smart Grids against PMU Adversarial Attacks: A Deep Learning-Based Robust Data Engineering Approach. Electronics 2023, 12, 2554. https://doi.org/10.3390/electronics12122554