Mavswarm2 is the ROS2-compatible version of Mavswarm. It supports simulating heterogenous quadrotor swarms of more than 10 robots on a single desktop with physics. Tested for ROS2 Jazzy.
New in Mavswarm2 is fully-distributed trajectory optimization with collision avoidance using sequential convex programming!
All the functionality is self-contained, and does not depend on any other ros packages except for simulator_interfaces.
Functionalities:
- quadrotor control (same geometric controller as mavswarm)
- trajectory optimization
- receding horizon planning
- collision avoidance
Consider citing our work [1][2] if you find this code helpful for your publications.
The internal controller uses the Lee's geometric tracking controller [3] and it is tuned for two different quadrotor models out of the box.
Compared to Mavswarm1, Mavswarm2 has a simpler installation, and the code is more self-contained. Everything is in the header files, so easier to manage and export as a library..
| Drone stabilization | Letter formation |
|---|---|
 |
 |
| A swarm of 10 drones stabilizing from an upside-down initialization | 10 drones forming letter "U" |
Install the dependencies
Install Eigen, Armadillo and GNU Science Library (GSL) before you continue.
Clone the repositories
mkdir -p mavswarm2/src
cd mavswarm2/src
git clone -b ros2 git@github.com:malintha/simulator_interfaces.git
git clone -b ros2 git@github.com:malintha/multi_uav_simulator.git
Building the simulator
cd ../
colcon build
Running the simulator
source install/setup.bash
ros2 launch mavswarm2 quadrotor.launch
Formation Control Goal Publishing
The example script publish_words.py publishes locations to the drones such that it can visualize a word, one letter-at-a-time using a formation of 10 drones. It computes a bipartite matching to find the least distance matching drones, then issue collision-free trajectory waypoints to be tracked using the internal geometric controller in a decentralized manner.
python publish_words.py HELLO
[1] Our work based on this controller:
@inproceedings{fernando2019formation,
title={Formation control and navigation of a quadrotor swarm},
author={Fernando, Malintha and Liu, Lantao},
booktitle={2019 International Conference on Unmanned Aircraft Systems (ICUAS)},
pages={284--291},
year={2019},
organization={IEEE}
}
[2] Mean-Field flocking control of UAVs (Find the complete code at: https://github.com/malintha/mean_field_flocking):
@INPROCEEDINGS{9560899,
author={Fernando, Malintha},
booktitle={2021 IEEE International Conference on Robotics and Automation (ICRA)},
title={Online Flocking Control of UAVs with Mean-Field Approximation},
year={2021},
volume={},
number={},
pages={8977-8983},
doi={10.1109/ICRA48506.2021.9560899}
}
[3] Geometric tracking controller:
@inproceedings{lee2010geometric,
title={Geometric tracking control of a quadrotor UAV on SE (3)},
author={Lee, Taeyoung and Leok, Melvin and McClamroch, N Harris},
booktitle={49th IEEE conference on decision and control (CDC)},
pages={5420--5425},
year={2010},
organization={IEEE}
}