MPConstellation is a group project for ME 231A at UC Berkeley. The goal of the project is to demonstrate model-predictive control (MPC) for a constellation of satellites.
Code has been testd on Python 3.6 and 3.7.
You will need to install pyomo and the ipopt solver.
We used a conda environment:
conda config --add channels conda-forge
conda config --set channel_priority strict
conda install -c conda-forge pyomo
conda install -c conda-forge ipopt
If any of these commands produce errors, try updating conda:
conda upgrade -n base conda
You may need to remove the conda-forge channel to upgrade, as it can prevent the environment from being solved:
conda config --remove channels conda-forge
To run a specific test method, use for example:
python3 -m unittest test_simulator.TestSimulator.test_get_trajectory_ODE
For a whole test suite:
python3 test_satellite.py
For all test suites (current working directory must be the repo directory):
python3 -m unittest
Calling Simulator.save_to_csv() will save one trajectory CSV for each satellite in the most recent simulation. The filename includes a timestamp and the Satellite object's unqiue ID.
To visualize simulation CSV(s), either use the Python plotting functions or plot in MATLAB:
visualizer('all')
will plot all CSVs in directory, while
visualizer('trajectory_xxx_yyy.csv')
will plot a specific CSV.