Online learning for robust voltage control under uncertain grid topology

Christopher Yeh, Jing Yu, Yuanyuan Shi, Adam Wierman
California Institute of Technology and UC San Diego

This repo contains code for the following two papers:

Robust online voltage control with an unknown grid topology
C. Yeh, J. Yu, Y. Shi, and A. Wierman
ACM e-Energy 2022, Best paper award finalist
Paper | Video

Online learning for robust voltage control under uncertain grid topology
C. Yeh, J. Yu, Y. Shi, and A. Wierman
Under submission
Preprint

Getting started

The code and instructions in this repo were tested on an Amazon AWS EC2 m5ad.8xlarge (32 CPU cores, 128 GiB RAM) instance running Ubuntu 22.04 LTS.

Install packages

1. Install miniconda3. Once miniconda3 is installed, we recommended that you use the libmamba solver for faster conda dependency resolution:

   conda config --set solver libmamba

2. Install the voltctrl conda environment:

   conda env update -f env.yml --prune

3. Request a Mosek license (link). Upon receiving the license file (mosek.lic) in an email, create a folder ~/mosek and copy the license file into that folder.

Running voltage control experiments

The main two scripts are run.py, which simulates bus voltages under approximate linearized distribution grid dynamics, and run_nonlinear.py, which simulatves bus voltages with a nonlinear balanced AC single-phase model.

Data Files (in /data)

The original data files were provided by the authors of the following paper:

Guannan Qu and Na Li. 2020. Optimal Distributed Feedback Voltage Control under Limited Reactive Power. IEEE Transactions on Power Systems 35, 1 (Jan. 2020), 315–331. https://doi.org/10.1109/TPWRS.2019.2931685

The original data files ("orig_data.zip") are attached to the releases. These original data files have been processed into the following files, which are the main files relevant for our experiments. See the inspect_matlab_data.ipynb notebook for details.

PV.mat

• contains single key 'actual_PV_profile'
• float64 array, shape [14421]
• min: 0.0, max: ~13.4
• units: MW
• description: solar generation, measured every 6 seconds

aggr_p.mat

• contains single key 'p'
• float64 array, shape [14421]
• min: ~2.4, max: ~7.1
• units: MW
• description: active power load, measured every 6 seconds for 24h

aggr_q.mat

• contains single key 'q'
• float64 array, shape [14421]
• min: ~1.1, max: ~3.1
• units: MVar
• description: reactive power load, measured every 6 seconds for 24h

pq_fluc.mat

• contains single key 'pq_fluc'
• float64 array, shape [55, 2, 14421]
• units: MW for p, MVar for q
• for p, min: ~-0.9, max: ~3.7
• for q, min: ~-0.5, max: ~0.0
• description: active and reactive power injection at 55 buses, measured every 6 seconds for 24h
  • first column is p, second column is q
  • • means generation, - means load
  • p is net active power injection (solar generation - load)

SCE_56bus.mat

• contains single key 'case_mpc'
• description: a "MATPOWER" file
• mat['case_mpc'][0,0] has 4 "keys"
  • 'version': shape [1], type uint8
  • 'baseMVA': shape [1, 1], type uint8, reference voltage at root bus
  • 'bus': shape [56, 13], type float64
  • 'branch': shape [55, 13], type float64
  • 'gen': shape [1, 21], type int16

nonlinear_voltage_baseline.npy

• float64 array, shape [14421, 56]
• description: balanced AC nonlinear simulation voltages, generated by nonlinear_no_control.py
• each column is the voltage of a bus, with column 0 being bus 0 (the substation)
• units: p.u. voltage (multiply by 12 to get kV)

See the attachments in releases for Python .pkl files containing the results of running the various algorithms. These Pickle files are read by the various Jupyter notebooks in the notebooks folder for plotting and analysis.

Citation

Please cite our papers as follows, or use the BibTeX entries below.

C. Yeh, J. Yu, Y. Shi, and A. Wierman, "Robust online voltage control with an unknown grid topology," in Proceedings of the Thirteenth ACM International Conference on Future Energy Systems (e-Energy '22), Association for Computing Machinery, Jun. 2022, pp. 240–250, ISBN: 9781450393973. DOI: 10.1145/3538637.3538853. [Online]. Available: https://dl.acm.org/doi/10.1145/3538637.3538853.

C. Yeh, J. Yu, Y. Shi, A. Wierman, "Online learning for robust voltage control under uncertain grid topology," Jun. 2023. DOI: 10.48550/arXiv.2306.16674. [Online]. Available: https://arxiv.org/abs/2306.16674.

@inproceedings{
    yeh2022robust,
    author = {Christopher Yeh and Jing Yu and Yuanyuan Shi and Adam Wierman},
    booktitle = {{Proceedings of the Thirteenth ACM International Conference on Future Energy Systems (e-Energy '22)}},
    doi = {10.1145/3538637.3538853},
    isbn = {9781450393973},
    month = {6},
    pages = {240-250},
    publisher = {Association for Computing Machinery},
    title = {Robust online voltage control with an unknown grid topology},
    url = {https://dl.acm.org/doi/10.1145/3538637.3538853},
    year = {2022}
}

@article{yeh2023online,
    author={Yeh, Christopher and Christianson, Nicolas and Low, Steven and Wierman, Adam and Yue, Yisong},
    month={6},
    title={{Online learning for robust voltage control under uncertain grid topology}},
    url={https://arxiv.org/abs/2306.16674},
    doi={10.48550/arXiv.2306.16674},
    year={2023}
}