Quick Start Guide

Once the installation of HEDGeOPF is successfully completed, it is possible to generate an AC-OPF dataset of N samples for a given power system test case by following few steps. Start by placing at a path of your choosing:

• the network data file in .m format (e.g., pglib_opf_case5_pjm.m)
• the configuration YAML file settings.yaml that controls the simulation.

Open the configuration YAML file and set entries grid and num_samples under section CASE to the grid file name as string (e.g., "pglib_opf_case5_pjm.m") and the desired number of AC-OPF samples N (e.g., 10000), respectively. If needed, change entry cpu_ratio under section PARALLEL to the desired percentage of CPU threads relative to Sys.CPU_THREADS that are used for distributed computing.

CASE:
    grid : "pglib_opf_case5_pjm.m"
    ...
    num_samples : 10000

...

PARALLEL:
    cpu_ratio : 50.0

Save and close the YAML file. Then, in Julia run

using HEDGeOPF
import HiGHS, Ipopt

path = "/path/of/your/choosing"
generate_dataset(path; filename = "settings.yaml")