Next Article in Journal
Redefining the MDA Framework—The Pursuit of a Game Design Ontology
Previous Article in Journal
Application of Pattern Language for Game Design in Pedagogy and Design Practice
Article

Detecting and Mitigating Adversarial Examples in Regression Tasks: A Photovoltaic Power Generation Forecasting Case Study

1
Department of Software Engineering, Pontifical Catholic University of Paraná, Londrina 86067000, Brazil
2
Department of Electrical Engineering, State University of Londrina, Londrina 86057970, Brazil
3
Department of Computer Science, State University of Londrina, Londrina 86057970, Brazil
*
Author to whom correspondence should be addressed.
Academic Editor: Willy Susilo
Received: 1 September 2021 / Revised: 17 September 2021 / Accepted: 18 September 2021 / Published: 26 September 2021
(This article belongs to the Section Artificial Intelligence)
With data collected by Internet of Things sensors, deep learning (DL) models can forecast the generation capacity of photovoltaic (PV) power plants. This functionality is especially relevant for PV power operators and users as PV plants exhibit irregular behavior related to environmental conditions. However, DL models are vulnerable to adversarial examples, which may lead to increased predictive error and wrong operational decisions. This work proposes a new scheme to detect adversarial examples and mitigate their impact on DL forecasting models. This approach is based on one-class classifiers and features extracted from the data inputted to the forecasting models. Tests were performed using data collected from a real-world PV power plant along with adversarial samples generated by the Fast Gradient Sign Method under multiple attack patterns and magnitudes. One-class Support Vector Machine and Local Outlier Factor were evaluated as detectors of attacks to Long-Short Term Memory and Temporal Convolutional Network forecasting models. According to the results, the proposed scheme showed a high capability of detecting adversarial samples with an average F1-score close to 90%. Moreover, the detection and mitigation approach strongly reduced the prediction error increase caused by adversarial samples. View Full-Text
Keywords: intelligent cyber-physical systems; adversarial machine learning; smart grid; security; photovoltaic generation forecast intelligent cyber-physical systems; adversarial machine learning; smart grid; security; photovoltaic generation forecast
Show Figures

Figure 1

MDPI and ACS Style

Santana, E.J.; Silva, R.P.; Zarpelão, B.B.; Barbon Junior, S. Detecting and Mitigating Adversarial Examples in Regression Tasks: A Photovoltaic Power Generation Forecasting Case Study. Information 2021, 12, 394. https://0-doi-org.brum.beds.ac.uk/10.3390/info12100394

AMA Style

Santana EJ, Silva RP, Zarpelão BB, Barbon Junior S. Detecting and Mitigating Adversarial Examples in Regression Tasks: A Photovoltaic Power Generation Forecasting Case Study. Information. 2021; 12(10):394. https://0-doi-org.brum.beds.ac.uk/10.3390/info12100394

Chicago/Turabian Style

Santana, Everton J., Ricardo P. Silva, Bruno B. Zarpelão, and Sylvio Barbon Junior. 2021. "Detecting and Mitigating Adversarial Examples in Regression Tasks: A Photovoltaic Power Generation Forecasting Case Study" Information 12, no. 10: 394. https://0-doi-org.brum.beds.ac.uk/10.3390/info12100394

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop