Docker Image with Project Toolsets

The various software tools required for this project are included in a docker image built by the Dockerfile given in roi-uncc/docker. This image is built by executing:

docker build -t roi-img .

in the directory Dockerfile exists in.

The image can be run using :

docker run --rm -v /ABSOLUTE/PATH/ON/HOST:/ABSOLUTE/PATH/ON/CONTAINER -i -t roi-img /bin/bash

This binds a local host directory to a directory in the container and is a good way to persist files between different invocations of this image.

The included open source softwares are: HELICS, GridLab-D and GridPACK.

To run the simulation

docker run --rm -it \
  -v "$PWD":/work \
  -w /work \
  roi-img \
  /bin/bash -lc 'rm -rf build && cmake -S . -B build && cmake --build build -j4 && helics run --path=ieee8500_gpk_cosim_runner.json'

For plot:

python plot_gpk_118.py --in gpk_118.csv --out_dir plots

To also plot the distribution-side voltage recorder output in the same run:

python plot_gpk_118.py --in gpk_118.csv --out_dir plots --dist-voltage sub_voltage.csv

If you also want a downstream node voltage plot:

python plot_gpk_118.py --in gpk_118.csv --out_dir plots --dist-voltage sub_voltage.csv --downstream-voltage downstream_voltage_A.csv

Modification:

• The feeder power is the summed power
• GridPACK solves the 118-bus case once per time step.
• The solved positive sequence bus voltage is republished to GridLAB-D as a balanced three phase voltage.
• Keep the unit dimension consistency while exchanging the date.
• Relocate the bus target. After partioning, relocate to bus number 2 to secure the right bus number.
• Form the iterative loop for convergence during the outer iteration. More specifically, the lines 17-90 in the powerflow_ex.cpp are added for the iterative loop.

One can dig into the aggregated_cosim_118 folder for more information.