This repository houses the code files for
Chapter 4 (Demonstrations and Applications)
of my master's thesis
(A Computational Approach to Anticipating Supply Injections and Bus Voltages in Steady-State Power Systems Analysis).
The complete repository
(which includes the manuscript and the data files)
is available at
https://doi.org/10.5281/zenodo.7523544.
The core anticipatory power flow (APF) routines are packaged as a small MATLAB library APFlib.
This library is built on MATPOWER
(specifically, version 7.1)
and uses CVX
(specifically, version 2.2 build 1148 (62bfcca)).
For a complete list of the dependencies, please see Table 4.1 of the manuscript.
To "install" APFlib,
- make sure that MATPOWER version 7.1 and CVX version 2.2 are available; and
- with the current working directory at the root of this repository,
invoke
installAPFlibin the MATLAB Command Window to installAPFlib.
Code files are organized as follows.
- Those pertaining to Section 4.1.1 (Solving the extended economic dispatch subproblem)
are in
time-apf1/. - Those pertaining to Section 4.1.2 (Solving the APF equations)
are in
time-apf2/. - Those pertaining to Section 4.2 (Effects of the supply regularization strengths)
are in
xed-regs/. - Those pertaining to Section 4.3.1 (Providing warm-start points)
are in
warm-start/. - Those pertaining to Section 4.3.2 (Finding the nearest power-flow feasible point)
are in
nearest-feasible/. - Those pertaining to Section 4.4 (Differentiating through the APF equations)
are in
diff-apf2/. - Those peratining to Section 4.5 (Probabilistic analysis)
are in
prob-analysis/. - Those pertaining to Appendix B.2 (The snapshot data sets)
are in
snapshots/.
All experiments and examples are implemented as MATLAB scripts (i.e., *.m files).
Snapshot data and raw results are available as .mat files
but are not included in this repository due to their size;
they can be downloaded from the versioned repository at
https://doi.org/10.5281/zenodo.7523544.
The processing of results (e.g., producing the plots used in the manuscript)
are implemented as Python scripts.
To run these processing scripts,
instantiate a Conda environment
using the file env.yml.
Please refer to the Conda user guide for more details.
We request that publications based on or derived from this repository,
or any part of it (e.g., the APFlib library) or of my thesis,
explicitly acknowledge that fact by citing my work.
Please use (also available in CITATION.bib):
@MastersThesis{CahigMastersThesis,
author = {Christian Y. {Cahig}},
date = {2022},
institution = {Mindanao State University - Iligan Institute of Technology},
title = {A Computational Approach to Anticipating Supply Injections and Bus Voltages in Steady-State Power Systems Analysis},
doi = {10.5281/zenodo.7523544},
address = {Iligan, Philippines},
}This repository is licensed under the Creative Commons Attribution 4.0 International Public License.
Please see LICENSE for the details.
This work is supervised jointly by Dr. Marven E. Jabian and Dr. Abdul Aziz G. Mabaning, and has been supported by the DOST-SEI ERDT program.