ModelicaGridData: Massive Power System Simulation Data Generation and Labeling Tool Using Modelica and Python

Sergio A. Dorado-Rojas, Fernando Fachini, Tetiana Bogodorova, Giuseppe Laera, Marcelo de Castro Fernandes, and Luigi Vanfretti

• Author's copy

• URL

bib

@article{dorado-rojas2023b,
  title = {{{ModelicaGridData}}: {{Massive Power System Simulation Data Generation}} and {{Labeling Tool Using Modelica}} and {{Python}}},
  shorttitle = {{{ModelicaGridData}}},
  author = {{Dorado-Rojas}, Sergio A. and Fachini, Fernando and Bogodorova, Tetiana and Laera, Giuseppe and {de Castro Fernandes}, Marcelo and Vanfretti, Luigi},
  year = 2023,
  month = feb,
  journal = {SoftwareX},
  volume = {21},
  issn = {23527110},
  doi = {10.1016/j.softx.2022.101258},
}

Abstract

This paper describes ModelicaGridData tool that is created for massive data generation employing phasor time-domain Modelica simulations and using the Open-Instance Power System Library (OpenIPSL). ModelicaGridData provides a pipeline for generating large amounts of data, considering a wide range of operating conditions and potential contingencies experienced by a power system. The need for large-scale power system dynamic data arises with the development of Machine Learning (ML) solutions in the context of the modernization of the existing power grid. ModelicaGridData implements algorithms to process different types of input data, perform steady-state computations, run dynamic simulations and linear analysis routines, and label the resulting data sets. The tool has been developed entirely in Python 3 and is compatible with the Modelica IDEs - Dymola and OpenModelica.

Back