[FEATURE] PGM Meet-up 2025-05-21 hackathon: validation cases from publicly available datasets

[nitbharambe]

Validation cases from publicly available datasets

This issue will guide you through the contribution of an IEEE test grid in PGM format to validate the accuracy and quality of calculations in the power-grid-model.

Premise

The modelling of various components and the algorithms for various calculations within power-grid-model are validated against established power system analysis tools and hand calculations. This ensures theoretical accuracy and reliability of the developed features. Power-grid-model (PGM) includes over 80 validation cases that are utilized to validate every incremental development of PGM. While all features are validated with minimal component examples, various practical or research focused larger grids can also be tested.

Since power-flow analysis results cannot be directly obtained from the inputs, these validation cases enhance confidence in the results produced by the PGM tool. The validation cases consist of a test input, an expected output, and a configuration file for the calculation. This task involves:

• Modelling publicly available grids in PGM using available IEEE datasets of PES test feeders.
• Converting the solutions to the PGM output format, which serves as the expected output for the calculation.
• Creating the validation case and contributing to the PGM repository.
• Verifying if the results match.

Steps

• Install power-grid-model along with development dependencies in Python. Check parent issue [Feature] PGM Meet-up 2025-05-21 Hackathons #977 for instructions on setup of the development environment. For this issue, the setup regarding contribution to PGM and python API is needed.
• (Optional) Create a script to generate data directly from the table provided below.
• Download the case data from [IEEE PES Test Feeders] (https://cmte.ieee.org/pes-testfeeders/resources/).
• Develop an input data model for PGM using the test cases, which should include topology, parameters, loads, and generation information from the feeder data.
• Specify any mutations to the input data in the update data section, such as load profiles for timeseries simulations.
• Create the expected output for the previously generated input/update data by referring to the solutions section.
  • You may choose to validate only specific results. For instance, if power results are adequately validated, voltage results may be omitted at your discretion (or vice-versa).
• Define the configuration parameters in params.json.
• Execute the validation tests.

Optional steps for creating script to create data from tables:

• Read the tables.
• Use initialize_array to fill the input PGM array. Check Components — power-grid-model documentation for details on all attributes.
• Fill the update PGM array.
• (Optional) Run PGM data validator: assert_valid_input_data to identify data errors in input.
• Fill the output PGM array.
• Export to a validation test case (TODO).