Skip to content

Latest commit

 

History

History
73 lines (53 loc) · 3.26 KB

README.md

File metadata and controls

73 lines (53 loc) · 3.26 KB
Raw

Terrestrial-Aerial-Navigation

Terrestrial-Aerial-Navigation is an autonomous navigation framework that brings complete autonomy to terrestrial-aerial bimodal vehicles (TABVs). This repository contains the following sub-modules:

  • A bi-level motion planner which generates safe, smooth, and dynamically feasible terrestrial-aerial hybrid trajectories.
  • A customized TABV platform that carries adequate sensing and computing resources while ensuring portability and maneuverability.

About

If our source code or hardware platform is used in your academic projects, please cite the related paper below.

@ARTICLE{Zhang2022TABV,
      author={Zhang, Ruibin and Wu, Yuze and Zhang, Lixian and Xu, Chao and Gao, Fei},
      journal={IEEE Robotics and Automation Letters}, 
      title={Autonomous and Adaptive Navigation for Terrestrial-Aerial Bimodal Vehicles}, 
      year={2022},
      volume={7},
      number={2},
      pages={3008-3015}
}

Video Links: Youtube.

Quick Start

Compiling tests passed on ubuntu 20.04. You can just execute the following commands one by one.

  1. Install nlopt following the official document.

  2. Install other dependencies and compile the project.

sudo apt-get install libarmadillo-dev
git clone https://github.com/ZJU-FAST-Lab/Terrestrial-Aerial-Navigation.git
cd Terrestrial-Aerial-Navigation
catkin_make
source devel/setup.bash
sh src/run.sh

Then, you can trigger the planner and choose the planning goal using the 2D Nav Goal tool in rviz. Then, the TABV will follow terrestrial-aerial hybrid trajectories to navigate a random forest map and cross a high barrier :

[NOTE] remember to change the CUDA option of src/uav_simulator/local_sensing/CMakeLists.txt, i.e., change the 'arch' and 'code' flags in the line of

set(CUDA_NVCC_FLAGS 
  -gencode arch=compute_75,code=sm_75;
)

according to your Nvidia graphics card version. You can check the right code here.

Acknowledgements

We build on Fast-Planner by extending its path searching and trajectory generation methods to TABV motion planning. We use NLopt for solving nonlinear optimization problems in trajectory generation.

Licence

The source code is released under GPLv3 license.

Maintenance

We are still working on extending the proposed system and improving code reliability.

For any technical issues, please contact Ruibin Zhang ([email protected]) or Fei Gao ([email protected]).

For commercial inquiries, please contact Fei Gao ([email protected]).