Corridor MPC: Towards Optimal and Safe Trajectory Tracking

We present a framework for safe and optimal trajectory tracking by combining Model Predictive Control and Sampled-Data Control Barrier functions. This framework, which we call Corridor MPC, safely and robustly keeps the state of the system within a corridor that is defined as a permissible error around a reference trajectory. By incorporating SampledData Control Barrier functions into an MPC framework, we guarantee safety for the continuous-time system in the sense of staying within the corridor and practical stability in the sense of converging to the reference trajectory

Running the demo

To install the package, make sure to have installed Python Poetry, Python 3.8 or 3.9 for CasADi compatibility, and python3-tk (for your Python version) to display plots. Then, run the following commands:

1. Clone the repository:

git clone git@github.com:KTH-DHSG/corridor_mpc.git

Optional: Configure poetry to create a local virtual environment:

poetry config virtualenvs.in-project true

2. Install the corridor_mpc library:

cd corridor_mpc
poetry install
poetry run python scripts/install_dependencies.py

NOTE: the script/install_dependencies will be successful in Ubuntu systems versions 18.04 or higher. For a different OS or version, please install the dependencies manually.

3. Run the demo script:

poetry run python scripts/run_tracking.py

Citing this work

To cite this work, please use the following BibTeX entry:

@inproceedings{9867764,
  author = {Roque, P. and Cortez, W. Shaw and Lindemann, L. and Dimarogonas, D. V.},
  booktitle = {2022 American Control Conference (ACC)},
  title = {Corridor MPC: Towards Optimal and Safe Trajectory Tracking},
  year = {2022},
  pages = {2025-2032},
  doi = {10.23919/ACC53348.2022.9867764},
}