Physics-Constrained Robustness Enhancement for Tree Ensembles Applied in Smart Grid
Source abstract
Abstract
With the widespread use of machine learning (ML) technology, the operational efficiency and responsiveness of power grids have been significantly enhanced, allowing smart grids to achieve high levels of automation and intelligence. However, tree ensemble models commonly used in smart grids are vulnerable to adversarial attacks, making it urgent to enhance their robustness. To address this, we propose a robustness enhancement method that incorporates physical constraints into the node-splitting decisions of tree ensembles. Our algorithm improves robustness by developing a dataset of adversarial examples that comply with physical laws, ensuring training data accurately reflects possible attack scenarios while adhering to physical rules. In our experiments, the proposed method increased robustness against adversarial attacks by 100% when applied to real grid data under physical constraints. These results highlight the advantages of our method in maintaining efficient and secure operation of smart grids under adversarial conditions.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Adversarial Robustness in Machine Learning · Smart Grid Security and Resilience · Network Security and Intrusion Detection
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (35)
Show source-provided references
1. Cyber physical system (CPS)-based Industry 4.0: A survey
2. Machine-to-machine communications with in-network data aggregation, processing, and actuation for large-scale cyber-physical systems10.1109/JIOT.2014.2311693 ↗
3. Artificial neural-net based dynamic security assessment for electric power systems10.1109/59.32481 ↗
4. Machine learning based power grid out age prediction in response to extreme events10.1109/TPWRS.2016.2631895 ↗
5. Reinforcement learning for demand response: A review of algorithms and modeling techniques10.1016/j.apenergy.2018.11.002 ↗
6. Modeling realistic adversarial attacks against network intrusion detection systems10.1145/3469659 ↗
7. Adversarial black-box attacks against network intrusion detection systems: A survey
8. An IoT-enabled stochastic operation management framework for smart grids10.1109/TITS.2022.3183327 ↗
9. A multi-layer security scheme for mitigating smart grid vulnerability against faults and cyber-attacks10.3390/app11219972 ↗
10. Cyber attacks on protective relays in digital substations and impact analysis
11. Cyber attacks on power system automation and protection and impact analysis
12. Threat analysis of blackenergy malware for synchrophasor based real-time control and monitoring in smart grid
13. P. Ken et al., “Sandworm Disrupts power in ukraine using a novel attack against operational technology,” 2023. Accessed: Jan. 20, 2024. [Online]. Available: https://cloud.google.com/blog/topics/threat-intelligence/sandworm-disrupts-power-ukraine-operational-technology/
14. Israeli hackers cause major disruptions in iranian electricity
15. On feasibility of coordinated time-delay and false data injection attacks on cyber-physical systems10.1109/JIOT.2021.3118065 ↗
16. Physics-constrained robustness evaluation of intelligent security assessment for power systems10.1109/TPWRS.2022.3169139 ↗
17. Transient stability assessment based on gated graph neural network with imbalanced data in internet of energy10.1109/JIOT.2021.3127895 ↗
18. Morison · Review of On-Line Dynamic Security Assessment Tools and Techniques · 2007
19. Online security assessment with load and renewable generation uncertainty: The itesla project approach
20. A systematic approach for dynamic security assessment and the corresponding preventive control scheme based on decision trees10.1109/TPWRS.2013.2283064 ↗
21. Towards deep learning models resistant to adversarial attacks
22. Evasion and hardening of tree ensemble classifiers
23. Intriguing properties of neural networks
24. Robust decision trees against adversarial examples
25. Efficient training of robust decision trees against adversarial examples
26. Cost-Aware robust tree ensembles for security applications
27. Attack on grid event cause analysis: An adversarial machine learning approach
28. Adaptive normalized attacks for learning adversarial attacks and defenses in power systems
29. Adversarial attacks and defense for CNN based power quality recognition in smart grid10.1109/TNSE.2021.3135565 ↗
30. Vulnerability of machine learning approaches applied in IoT-based smart grid: A review10.1109/JIOT.2024.3349381 ↗
31. Vulnerability analysis, robustness verification, and mitigation strategy for machine learning-based power system stability assessment model under adversarial examples10.1109/TSG.2021.3133604 ↗
32. Resilience enhancement of multi-agent reinforcement learning-based demand response against adversarial attacks10.1016/j.apenergy.2022.119688 ↗
33. Data security of machine learning applied in low-carbon smart grid: A formal model for the physics-constrained robustness10.1016/j.apenergy.2023.121405 ↗
34. Physics-constrained vulnerability assessment of deep reinforcement learning-based SCOPF10.1109/TPWRS.2022.3192558 ↗
35. MATPOWER: Steady-state operations, planning, and analysis tools for power systems research and education10.1109/TPWRS.2010.2051168 ↗