Python library for simulating dynamics and ensemble model predictive control.

The code in this repository was prepared to implement the methodology described in

1. C. Folkestad, D. Pastor, J. Burdick, "Ensemble Model Predictive Control", in Proc. Conf on Decision and Control Control, (submitted) 2020

The simulation framework of this repository is adapted from the Learning and Control Core Library.

Set up virtual environment

python3 -m venv .venv

Activate virtual environment

source .venv/bin/activate

Upgrade package installer for Python

pip install --upgrade pip

Install requirements

pip3 install -r requirements.txt

Create conda environment

conda create --name ensemblempc
conda activate ensemblempc

conda install matplotlib numpy pyqtgraph
pip install torch cvxpy % not available in conda

To run the code, run one of the examples in

core/examples

Run the example scripts as a module with the root folder of repository as the working directory. For example, in a Python 3 environment run

python -m core.examples.1d_drone_landing

To visualize use

python -m core.examples.1d_pyqtgraph