This code accompanies the RA-L/ICRA 2020 paper
C. E. Luis, M. Vukosavljev, and A. P. Schoellig, “Online trajectory generation with distributed model predictive control for multi-robot motion planning,” IEEE Robot. Autom. Lett., vol. 5, no. 2, pp. 604–611, Jan. 2020.
If you use this library for your own work, consider citing:
@article{luis2020online,
title={Online trajectory generation with distributed model predictive control for multi-robot motion planning},
author={Luis, Carlos E and Vukosavljev, Marijan and Schoellig, Angela P},
journal={IEEE Robotics and Automation Letters},
volume={5},
number={2},
pages={604--611},
year={2020},
publisher={IEEE}
}
- Standalone C++ library implementing the algorithm.
- MATLAB code for running the benchmark and visualize data.
Below you will find instructions on how to use the two main pieces of software included in this repo.
Dependencies:
- C++14
- Cmake >= 3.0
- Eigen >= 3.0
- Initialize qpOASES submodule
cd <path-to-repo>
git submodule init && git submodule update
- Build the project
mkdir build
cd build
cmake ..
make
- Test installation by running example scenario
cd ../bin
./run
- You should see a stream of data in the console. The expected last lines of the console output are:
No collisions found!
All the vehicles reached their goals!
Writing solution to text file...
- The generated simulation data is in
cpp/results/trajectories.txt
. You can run the MATLAB scriptplot_results.m
for a 3D visualization of the generated trajectories.
The entry point of the code is src/main.cpp
. If you want to run your own transition scenarios, or play around with the (many) hyperparameters of the algorithm, the main configuration file is in cpp/config/config.json
. You can find an explanation of each hyperparameter in cpp/config/help.txt
.
The code was written and executed with MATLAB2018a. There's no guarantees it will work in other versions.
To run the benchmark presented in the paper against the Buffered Voronoi Cells method, execute matlab/tests/comp_dmpc_bvc.m
. At the top of the file there's several parameters to change the test characteristics.
plot_2agent_video.m
: dynamic 3D visualization of experiment with 2 drones exchanging positionsplot_comp_allCA.m
: summarize results from benchmark against BVC methodplot_disturbance_exp.m
: plots in paper for continuous vs event-based replanningplot_hoop_test_paper.m
: static 3D visualization of 10 drones passing through a hula-hoopplot_hoop_test_video.m
: dynamic 3D visualization of 10 drones passing through a hula-hoopplot_obstaclefree_video.m
: dynamic 3D visualization of 20 drones randomly transitioning between goal points