This is a ROS workspace that creates a trajectory for a UAV to follow passing through a set of given waypoints and avoiding a set of given cylindrical obstacles, using a path planning algorithm. The testing is done through a node which plots the waypoints, obstacles and the current pose of UAV on RVIZ for examining the accuracy of the algorithm.
- ROS
- ardupilot
- mavros
- rviz
- Mission Planner (preferred) or apm planner
1. roscore
2. /(path to sim_vehicle)/sim_vehicle.py --console --map --aircraft test
3. roslaunch mavros apm2.launch fcu_url:=udp://localhost:14550@
4. rosrun map currentXY
5. rosrun map markPoints
6. rviz
(write the frame id i.e. /my_frame in the Fixed Frame)
7. rosrun tf static_transform_publisher 0 0 0 0 0 0 1 map my_frame 10
8. rosrun map waypoints
@misc{bhagat-obstacle-simulator-ros,
author = {Sarthak Bhagat},
title = {sarthak268/Obstacle_Avoidance_for_UAV},
url = {https://github.com/sarthak268/Obstacle_Avoidance_for_UAV},
year = {2018}
}
You may also want to look at the following paper (accepted at ICUAS'20).
@article{Bhagat2020UAVTT,
title={UAV Target Tracking in Urban Environments Using Deep Reinforcement Learning},
author={Sarthak Bhagat and P. B. Sujit},
journal={ArXiv},
year={2020},
volume={abs/2007.10934}
}
For any queries, please contact me via mail on sarthak16189@iiitd.ac.in
Like my work? Buy me a coffee: https://www.buymeacoffee.com/sarthakbhagat