HIGHSTAR: High-Speed and Efficient Online Autonomous UAV Exploration(TASE 2025 accepted)
Highstar is a high-speed and efficient online autonomous UAV exploration method. First, to fully leverage the current velocity and acceleration of the UAV in exploration target planning, we propose a motion primitive activated graph search method. Specifically, it utilizes motion primitives to simulate the motion tendency in a short period to improve the accuracy of evaluated motion time cost and searches from the terminal points of motion primitives on a voxel graph with a dynamically calculated searching upper bound to reduce the computational cost. Then, a minimum time trajectory to the optimal viewpoint with a non-zero terminal velocity constraint inside a safety-guaranteed and exploration-stimulative convex hull is optimized. Finally, a SE(3) coverage trajectory that can cover the unknown space around the exploration path is further optimized according to the minimum time trajectory.
If you found this research useful for your own work, please use the following citation:
@article{0HIGHSTAR,
title={HIGHSTAR: High-Speed and Efficient Online Autonomous UAV Exploration},
author={ Dong, Qianli and Zhang, Xuebo and Zhang, Shiyong and Wang, Ziyu and Ma, Zhe and Li, Tianyi and Xi, Haobo },
journal={IEEE Transactions on Automation Science and Engineering},
year={2025},
volume={22},
}
And this work is based on our previus work:
@article{2025FSMP,
title={FSMP: A Frontier-Sampling-Mixed Planner for Fast Autonomous Exploration of Complex and Large 3-D Environments},
author={ Zhang, Shiyong and Zhang, Xuebo and Dong, Qianli and Wang, Ziyu and Xi, Haobo and Yuan, Jing },
journal={IEEE Transactions on Instrumentation and Measurement},
year={2025},
volume={74},
}
Author: Qianli Dong (Charlie Dog), Shiyong Zhang, Ziyu Wang, Zhe Ma, Tianyi Li, and Haobo Xi.
This work is developed in Ubuntu 20.04, ROS noetic.
Prerequisites:
$ sudo apt-get install ros-noetic-joy ros-noetic-octomap-ros python3-wstool python3-catkin-tools protobuf-compiler libgoogle-glog-dev ros-noetic-control-toolbox ros-noetic-mavros
Simulation environment: We adopt RotorS as the simulation platform. To adapt to the scenario of multi-UAV exploration, we have made some modifications. You can get the modified version with:
$ mkdir -p rotors-ws/src
$ cd rotors-ws/src
$ git clone https://github.com/NKU-UAVTeam/rotors-modified.git
$ git clone https://github.com/ethz-asl/mav_comm.git
$ git clone https://github.com/catkin/catkin_simple.git
$ cd ..
$ catkin_make
# if report error: "‘mavlink_status_t’ has not been declared", $sudo apt remove ros-noetic-mavlink. install it later
$ echo "source ~/rotors-ws/devel/setup.bash --extend" >> ~/.bashrc
Clone Code and Make:
$ mkdir -p HighStar/src
$ cd HighStar/src
$ git clone https://github.com/ethz-asl/gflags_catkin.git
$ git clone https://github.com/ethz-asl/glog_catkin.git
$ git clone https://github.com/catkin/catkin_simple.git
$ git clone https://github.com/NKU-MobFly-Robotics/HighStar.git
$ cd ..
$ catkin_make
Run Exploration:
$ source devel/setup.bash
$ roslaunch murder murder_demo_maze4_mid.launch
Show Explored Volume:
rostopic echo /
Exploration Space: Parameters are contained in the yaml files: murder_maze4.yaml
# exploration bounding box
Exp/minX: -10.0
Exp/minY: -20.0
Exp/minZ: 0.0
Exp/maxX: 10.0
Exp/maxY: 20.0
Exp/maxZ: 3.0
# map size (larger than the exploration bounding box)
block_map/minX: -22.5
block_map/minY: -22.5
block_map/minZ: -0.1
block_map/maxX: 22.5
block_map/maxY: 22.5
block_map/maxZ: 3.5
We use MINCO for trajectory planning.
Thank you to @Killow1998 for pointing out these issues.
- Section IV-E, the paper states ''In high-speed trajectory planning, the tilt of the UAV is negligible ...'', here, the ''negligible'' should be ''non-negligible''.
- Section IV-E, before Eq. 18, ''Thus, the safety cost is...'', it should be ''feasibility cost''.
Maybe an aerobatic exploration planner will come soon. Maybe.