Ewok: 使用统一 B-样条和 3D 圆形缓冲区对 MAV 进行实时轨迹重规划
For more information see https://vision.in.tum.de/research/robotvision/replanning
If you use this code, please cite the respective publications:
The system has been tested with Ubuntu 20.04 (ROS Noetic) and Ubuntu 18.04 (ROS Melodic). For Ubuntu 16.04 (ROS Kinetic) switch to ubuntu_16_04_kinetic branch.
Follow the tutorials to install ROS Noetic for 20.04 or ROS Melodic for 18.04 (desktop-full) and to set up catkin workspace.
Install dependencies. For Ubuntu 20.04 (ROS Noetic):
sudo apt-get install git python python3-matplotlib python3-numpy libeigen3-dev libgoogle-glog-dev libatlas-base-dev libsuitesparse-dev protobuf-compiler libnlopt-dev libnlopt-cxx-dev ros-noetic-octomap ros-noetic-octomap-ros ros-noetic-octomap-msgs ros-noetic-tf-conversions ros-noetic-eigen-conversionsFor Ubuntu 18.04 (ROS Melodic):
sudo apt-get install git python3-matplotlib python3-numpy libeigen3-dev libgoogle-glog-dev libatlas-base-dev libsuitesparse-dev protobuf-compiler libnlopt-dev ros-melodic-octomap ros-melodic-octomap-ros ros-melodic-octomap-msgs ros-melodic-tf-conversions ros-melodic-eigen-conversionsNavigate to the source folder of your catkin workspace, download and build the software:
cd ~/catkin_ws/src
git clone --recursive https://github.com/VladyslavUsenko/ewok.git
cd ..
catkin_makeSome of the provided examples (trajectory_replanning_big_forest.launch) depend on the forest_gen dataset. With the above commands it should already be downloaded as a git submodule. Please check https://github.com/ethz-asl/forest_gen for more information about the dataset.
In separate terminal windows:
Start roscore:
roscoreStart visualization:
roscd ewok_simulation/rviz/
rviz -d simulation.rvizLaunch the system and simulator:
roslaunch ewok_simulation trajectory_replanning_simulation.launchNow you should be able to see Rviz visualization of the system running in simulator. Due to performance reasons GUI of the Gazebo simulator is disabled by default.
To enable it, change the following parameter in ewok_simulation/launch/trajectory_replanning_simulation.launch:
In separate terminal windows:
Start roscore:
roscoreor, if you already have it running, make sure you are not using simulated time:
rosparam set use_sim_time falseStart visualization:
roscd ewok_ring_buffer/rviz/
rviz -d ring_buffer.rvizRun benchmarking script. It will download the dataset, run the computations and show the plots with insertion times:
roscd ewok_ring_buffer/benchmarking/
./benchmark.pyIn separate terminal windows:
Start roscore:
roscoreor, if you already have it running, make sure you are not using simulated time:
rosparam set use_sim_time falseStart visualization:
roscd ewok_optimization/rviz/
rviz -d optimization_example.rvizRun optimization example:
rosrun ewok_optimization spline_optimization_example3Ewok was developed at the Technical University of Munich. The open-source version is licensed under the GNU Lesser General Public License Version 3 (LGPLv3).
暂无开放 Issues,或尚未同步最近议题。