mpclab_strategy_obca

Collision Avoidance in Tightly-Constrained Environments without Coordination: a Hierarchical Control Approach

Webpage, arXiv

Authors:

• Xu Shen, [email protected]
• Edward Zhu, [email protected]
• Yvonne R. Stürz, [email protected]

A ROS and python package containing the code for online receding horizon OBCA control in tightly-constrained environments leveraging data-driven strategy prediction. The MATLAB simultion can be found in repo dataMPC_parking.

This package is intended for use with the BARC_research simulation and experiment codebase.

To use the controllers in this package, navigate to the root directory of this repository and run catkin_make. The setup.bash file from this package should be sourced after the one from BARC_research using the following command from the root directory

source ./devel/setup.bash --extend

The --extend flag stops overwriting from sourcing the setup.bash from this package.

Dependencies:

• APT: libgmp3-dev (required for pypoman)
• Python: numpy, scipy, torch, pypoman, pyclipper, casadi==3.5.1, forcespro