ABSTRACT
Recent years have seen a rich literature of data-driven approaches designed for power grid applications. However, insufficient consideration of domain knowledge can impose a high risk to the practicality of the methods. Specifically, ignoring the grid-specific spatiotemporal patterns (in load, generation, and topology, etc.) can lead to outputting infeasible, unrealizable, or completely meaningless predictions on new inputs. To address this concern, this paper investigates real-world operational data to provide insights into power grid behavioral patterns, including the time-varying topology, load, and generation, as well as the spatial differences (in peak hours, diverse styles) between individual loads and generations. Then based on these observations, we evaluate the generalization risks in some existing ML works caused by ignoring these grid-specific patterns in model design and training.
- Defense Use Case. 2016. Analysis of the cyber attack on the Ukrainian power grid. Electricity Information Sharing and Analysis Center (E-ISAC) 388 (2016).Google Scholar
- Priya L Donti, David Rolnick, and J Zico Kolter. 2021. DC3: A learning method for optimization with hard constraints. arXiv preprint arXiv:2104.12225 (2021).Google Scholar
- Bryan Hooi, Dhivya Eswaran, Hyun Ah Song, Amritanshu Pandey, Marko Jereminov, Larry Pileggi, and Christos Faloutsos. 2018. GridWatch: Sensor Placement and Anomaly Detection in the Electrical Grid. In ECML-PKDD. Springer, 71–86.Google Scholar
- Xinyue Hu, Haoji Hu, Saurabh Verma, and Zhi-Li Zhang. 2020. Physics-guided deep neural networks for power flow analysis. IEEE Transactions on Power Systems 36, 3 (2020), 2082–2092.Google ScholarCross Ref
- Ethan King, Ján Drgoňa, Aaron Tuor, Shrirang Abhyankar, Craig Bakker, Arnab Bhattacharya, and Draguna Vrabie. 2022. Koopman-based Differentiable Predictive Control for the Dynamics-Aware Economic Dispatch Problem. In 2022 American Control Conference (ACC). 2194–2201. https://doi.org/10.23919/ACC53348.2022.9867379Google Scholar
- James Kotary, Ferdinando Fioretto, Pascal Van Hentenryck, and Bryan Wilder. 2021. End-to-end constrained optimization learning: A survey. arXiv preprint arXiv:2103.16378 (2021).Google Scholar
- Ognjen Kundacina, Mirsad Cosovic, and Dejan Vukobratovic. 2022. State Estimation in Electric Power Systems Leveraging Graph Neural Networks. arXiv preprint arXiv:2201.04056 (2022).Google Scholar
- T. Kërçi, J. Giraldo, and F. Milano. 2020. Analysis of the impact of sub-hourly unit commitment on power system dynamics. International Journal of Electrical Power & Energy Systems 119 (2020), 105819. https://doi.org/10.1016/j.ijepes.2020.105819Google ScholarCross Ref
- Shimiao Li, Jan Drgona, Aaron R Tuor, Larry Pileggi, and Draguna L Vrabie. [n. d.]. Homotopy Learning of Parametric Solutions to Constrained Optimization Problems. ([n. d.]).Google Scholar
- Shimiao Li, Amritanshu Pandey, Bryan Hooi, Christos Faloutsos, and Larry Pileggi. 2021. Dynamic graph-based anomaly detection in the electrical grid. IEEE Transactions on Power Systems 37, 5 (2021), 3408–3422.Google ScholarCross Ref
- Shimiao Li, Amritanshu Pandey, and Larry Pileggi. 2022. GridWarm: Towards Practical Physics-Informed ML Design and Evaluation for Power Grid. arXiv preprint arXiv:2205.03673 (2022).Google Scholar
- Shimiao Li, Amritanshu Pandey, and Larry Pileggi. 2023. Contingency Analyses with Warm Starter using Probabilistic Graphical Model. arxiv:2304.06727 [cs.CR]Google Scholar
- Gaoqi Liang, Junhua Zhao, Fengji Luo, Steven R Weller, and Zhao Yang Dong. 2016. A review of false data injection attacks against modern power systems. IEEE Transactions on Smart Grid 8, 4 (2016), 1630–1638.Google ScholarCross Ref
- Mostafa Mohammadian, Kyri Baker, and Ferdinando Fioretto. 2022. Gradient-enhanced physics-informed neural networks for power systems operational support. arXiv preprint arXiv:2206.10579 (2022).Google Scholar
- Juan Ospina, Xiaorui Liu, Charalambos Konstantinou, and Yury Dvorkin. 2020. On the feasibility of load-changing attacks in power systems during the covid-19 pandemic. IEEE Access 9 (2020), 2545–2563.Google ScholarCross Ref
- PJM. 2022. PJM Manual 03: Transmission Operations. Revision: 63. uments/manuals/m03.ashxGoogle Scholar
- PJM. 2023. PJM Manual 11: Energy and Ancillary Services Market Operations. Revision: 123. https://www.pjm.com/ /media/documents/manuals/m11.ashxGoogle Scholar
- Brian Singer, Amritanshu Pandey, Shimiao Li, Lujo Bauer, Craig Miller, Lawrence Pileggi, and Vyas Sekar. 2022. Shedding Light on Inconsistencies in Grid Cybersecurity: Disconnects and Recommendations. In 2023 IEEE Symposium on Security and Privacy (SP). IEEE Computer Society, 554–571.Google Scholar
- Narri Yadaiah and G Sowmya. 2006. Neural network based state estimation of dynamical systems. In The 2006 IEEE international joint conference on neural network proceedings. IEEE, 1042–1049.Google Scholar
- Ahmed S Zamzam and Kyri Baker. 2020. Learning optimal solutions for extremely fast AC optimal power flow. In 2020 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm). IEEE, 1–6.Google ScholarCross Ref
Recommendations
Risk assessment method for power grid communication link failure based on fuzzy comprehensive assessment
In order to overcome the problems of long time and inaccurate evaluation results of traditional power grid communication link fault assessment methods, a new power grid communication link fault risk assessment algorithm based on fuzzy comprehensive ...
Power Grid Vulnerability Identifying Based on Complex Network Theory
IMCCC '12: Proceedings of the 2012 Second International Conference on Instrumentation, Measurement, Computer, Communication and ControlThe frequent occurrence of power grid blackouts draws more and more attention on the mechanism of cascading failure and the vulnerability of power grid. In order to analyze the problem existed in the power grid and provide useful reference for power ...
Indices System and Methods for Power Grid Security Assessment Based on Dynamic Fuzzy
ICECE '10: Proceedings of the 2010 International Conference on Electrical and Control EngineeringThe power grid security assessment is used to analyse the existing dangerous and harmful factors of power grid, judging the potential risks as well. The assessment results can provide scientific basis for prevention of accidents and guide management ...
Comments