Copyright (c) 2024 Alliance for Sustainable Energy, LLC.
NREL software record: SWR-24-59
Numba-compatible self-adaptive differential evolution algorithm for optimization tasks. This implementation is based on the method introduced by Brest, Janez, et al., in their study "Self-adapting control parameters in differential evolution: A comparative study on numerical benchmark problems," published in IEEE Transactions on Evolutionary Computation, Vol. 10, No. 6, 2006, pp. 646-657.
Script: Facilitates the extraction of single diode parameters from IV data, streamlining the analysis process.
Script: Exemplifies the application of the DE algorithm to minimize the 2D Ackley function
Script: Results from minimizing the Ackley function
Jupyter Notebook: Employed for generating figures based on the data collected, aiding in the visualization and interpretation of results.
Method Evaluation: Evaluates the main branch of the Lambert W function, approximated via the Newton-Raphson method, which is then compared against results from the SciPy implementation.
If you use this repository in your research or find it helpful, please cite the following paper:
@article{febba2025_dd,
author = "Fébba, Davi and Egbo, Kingsley and Callahan, William A. and Zakutayev, Andriy",
title = "From text to test: AI-generated control software for materials science instruments",
journal = "Digital Discovery",
year = "2025",
volume = "4",
issue = "1",
pages = "35-45",
publisher = "RSC",
doi = "10.1039/D4DD00143E",
url = "http://dx.doi.org/10.1039/D4DD00143E"
}