Predicting AC Optimal Power Flows: Combining Deep Learning and Lagrangian Dual Methods
Open accessAbstract
The Optimal Power Flow (OPF) problem is a fundamental building block for the optimization of electrical power systems. It is nonlinear and nonconvex and computes the generator setpoints for power and voltage, given a set of load demands. It is often solved repeatedly under various conditions, either in real-time or in large-scale studies. This need is further exacerbated by the increasing stochasticity of power systems due to renewable energy sources in front and behind the meter. To address these challenges, this paper presents a deep learning approach to the OPF. The learning model exploits the information available in the similar states of the system (which is commonly available in practical applications), as well as a dual Lagrangian method to satisfy the physical and engineering constraints present in the OPF. The proposed model is evaluated on a large collection of realistic medium-sized power systems. The experimental results show that its predictions are highly accurate with average errors as low as 0.2%. Additionally, the proposed approach is shown to improve the accuracy of the widely adopted linear DC approximation by at least two orders of magnitude.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
ML can be used to approximate or simplify existing optimization problems, find good starting points for optimization, identify redundant constraints, learn from the actions of power system control engineers, or do some combination of these.
“ML can be used to approximate or simplify existing optimization problems [75,242,311,871], find good starting points for optimization [52,196,382], identify redundant constraints [541], learn from the actions of power system control engineers [197], or do some combination of these [858].”
From Tackling climate change with machine learning (2023) · cites this paper as [242]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126 · see citing works
(2017) · Cited by 11,625 · see citing works
Wächter, A., & Biegler, L. T. (2006) · Mathematical Programming · doi:10.1007/s10107-004-0559-y · Cited by 9,483 · see citing works
Power Generation, Operation, and Control
Wood, A. J., & Wollenberg, B. F. (1984) · Cited by 8,337 · see citing works
Multiplier and gradient methods
Hestenes, M. R. (1969) · Journal of Optimization Theory and Applications · doi:10.1007/bf00927673 · Cited by 2,507 · see citing works
Optimal capacitor placement on radial distribution systems
Baran, M. E., & Wu, F. F. (1989) · IEEE Transactions on Power Delivery · doi:10.1109/61.19265 · Cited by 2,062 · see citing works
Real-Time Motor Fault Detection by 1-D Convolutional Neural Networks
İnce, T., Serkan Kiranyaz, B., Eren, L., Aşkar, M., & Gabbouj, M. (2016) · IEEE Transactions on Industrial Electronics · doi:10.1109/tie.2016.2582729 · Cited by 1,419 · see citing works
Security-Constrained Unit Commitment With Volatile Wind Power Generation
Wang, J., Shahidehpour, M., & Li, Z. (2008) · IEEE Transactions on Power Systems · doi:10.1109/tpwrs.2008.926719 · Cited by 933 · see citing works
A review of recent advances in economic dispatch
Chowdhury, B., & Rahman, S. (1990) · IEEE Transactions on Power Systems · doi:10.1109/59.99376 · Cited by 686 · see citing works
Optimal Transmission Switching
Fisher, E. B., O’Neill, R. P., & Ferris, M. C. (2008) · IEEE Transactions on Power Systems · doi:10.1109/tpwrs.2008.922256 · Cited by 621 · see citing works
Security-Constrained Optimal Power Flow with Post-Contingency Corrective Rescheduling
Monticelli, A., Pereira, M. V. F., & Granville, S. (1987) · IEEE Transactions on Power Systems · doi:10.1109/tpwrs.1987.4335095 · Cited by 591 · see citing works
PowerModels. JL: An Open-Source Framework for Exploring Power Flow Formulations
Coffrin, C., Bent, R. W., Sundar, K., Ng, Y., & Lubin, M. (2018) · Power Systems Computation Conference · doi:10.23919/pscc.2018.8442948 · Cited by 286 · see citing works
Tang, Y., Dvijotham, K., & Low, S. H. (2017) · IEEE Transactions on Smart Grid · doi:10.1109/tsg.2017.2704922 · Cited by 253 · see citing works
NESTA, The NICTA Energy System Test Case ArchivePreprint
Coffrin, C., Gordon, D., & Scott, P. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1411.0359 · Cited by 127 · see citing works
Security-constrained unit commitment with volatile wind power generation
Wang, J., Shahidehpour, M., & Li, Z. (2009) · IEEE Power & Energy Society General Meeting (PESGM) · doi:10.1109/pes.2009.5275867 · Cited by 111 · see citing works
Learning for DC-OPF: Classifying active sets using neural nets
Deka, D., & Misra, S. (2019) · doi:10.1109/ptc.2019.8810819 · Cited by 101 · see citing works
Boukelia, T. E., Arslan, O., & Mecibah, M. S. (2016) · Renewable Energy · doi:10.1016/j.renene.2016.12.081 · Cited by 82 · see citing works
Statistical Learning for DC Optimal Power Flow
Ng, Y., Misra, S., Roald, L., & Backhaus, S. (2018) · Power Systems Computation Conference · doi:10.23919/pscc.2018.8442859 · Cited by 62 · see citing works
Deep Learning for Power System Security Assessment
Arteaga, J.-M. H., Hancharou, F., Thams, F., et al. (2019) · doi:10.1109/ptc.2019.8810906 · Cited by 56 · see citing works
Transmission expansion planning: A review
Niharika, N., Verma, S., & Mukherjee, V. (2016) · International Conference on Energy Efficient Technologies for Sustainability · doi:10.1109/iceets.2016.7583779 · Cited by 39 · see citing works
Anwar, M., Moursi, M. S. E., & Xiao, W. (2016) · IEEE Transactions on Power Systems · doi:10.1109/tpwrs.2016.2591723 · Cited by 34 · see citing works
Constraint-Based Lagrangian Relaxation
Fontaine, D., Michel, L. D., & Hentenryck, P. V. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-10428-7_25 · Cited by 20 · see citing works
Look-ahead multi-time frame generator control and dispatch method in PJM real time operations
Tong, J., & Ni, H. J. (2011) · IEEE Power and Energy Society General Meeting (PESGM) · doi:10.1109/pes.2011.6039726 · Cited by 18 · see citing works
Chatziagorakis, P., Ziogou, C., Elmasides, C., et al. (2016) · Neural Computing and Applications · doi:10.1007/s00521-015-2175-6 · Cited by 13 · see citing works
LEAP nets for power grid perturbationsPreprint
Donnot, B., Donon, B., Guyon, I., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1908.08314 · Cited by 10 · see citing works
Enhanced pan-European transmission planning methodology
Pache, C., Maeght, J., Seguinot, B., et al. (2015) · IEEE Power & Energy Society General Meeting (PESGM) · doi:10.1109/pesgm.2015.7286270 · Cited by 5 · see citing works
Flagged first, then by citations
Research works printed but not linked: 2
- [Fioretto, Mak, and Van Hentenryck 2020] Fioretto, F.; Mak, T. W. K.; and Van Hentenryck, P. 2020. Predicting AC optimal power flows: Combining deep learning and lagrangian dual methods. In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI) , to appear.
- [Liu et al. 2018] Liu, J.; Marecek, J.; Simonetta, A.; and Taka, M. 2018. A coordinate-descent algorithm for tracking solutions in time-varying optimal power flows. In Power Systems Computation Conference .
Cited by
Sorted by citations
Perspectives on the integration between first-principles and data-driven modeling
Bradley, W., Kim, J., Kilwein, Z., Blakely, L., Eydenberg, M., Jalvin, J., Laird, C., & Boukouvala, F. (2022) · Computers & Chemical Engineering · doi:10.1016/j.compchemeng.2022.107898 · Cited by 231 · see citing works
Tackling climate change with machine learning
Rolnick, D., Donti, P. L., Kaack, L. H., Kochanski, K., Lacoste, A., Sankaran, K., Ross, A. S., Milojevic-Dupont, N., Jaques, N., Waldman-Brown, A., et al. (2023) · ACM Computing Surveys · doi:10.1145/3485128 · Cited by 336 · see citing works
- Authors
- Fioretto, F., Mak, T. W. K., & Van Hentenryck, P.
- Affiliations
- Syracuse University (US)
- Georgia Institute of Technology (US)
- Venue
- Proceedings of the AAAI Conference on Artificial Intelligence
- Published
- 2020-04-03
- Volume
- 34
- Issue
- 01
- Pages
- 630-637
- ISSN
- 2159-5399 (ISSN-L)
- 2374-3468
- OpenAlex
- W2973782063
- Cited by
- 29
- References
- 26
- Citations by year
- 2025: 3 · 2024: 6 · 2023: 3 · 2022: 5 · 2021: 7 · 2020: 4 · 2019: 1
- FWCI
- 6.68
- Type
- conference paper
- Language
- en
- Open access
- diamond
- doi.org
- Licence
- www.aaai.org/ (unspecified)
- Locations
- Proceedings of the AAAI Conference on Artificial Intelligence (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Science Foundation (1709094)
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Optimal Power Flow Distribution
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Electric Power System Optimization
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Power System Optimization and Stability
- Keywords
- optimal power flow, AC power flow, deep learning, Lagrangian dual method, power system optimization, nonlinear nonconvex optimization, renewable energy integration, power system constraints
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
