8000 GitHub - xiaofeifei-1/HKUSTGZ_Indoor_Robot
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

xiaofeifei-1/HKUSTGZ_Indoor_Robot

BranchesTags
 
 

Repository files navigation

VIP_ROBOT

update on 17/11/2024

Related Repository

Complete the navigation function with lidar positioning and ros_navigation

navigation_config： Contains all the parameter files needed for ros_navigation.

fast_lio：https://github.com/hku-mars/FAST_LIO

fast_lio_sam_qn：https://github.com/engcang/FAST-LIO-SAM-QN

fast_lio_localization_qn : Base on https://github.com/engcang/FAST-LIO-Localization-QN

ranger_mini_v2 : Base on https://github.com/agilexrobotics/ugv_gazebo_sim, (Cut out the original huge package and keep only the parts that are relevant to ranger_mini_v2.)

ros environment : https://wiki.ros.org/noetic/Installation/Ubuntu (base on your device)

Videos

Testing Indoor

Prerequisites

  • C++ >= 17
  • OpenMP >= 4.5
  • CMake >= 3.10.0
  • Eigen >= 3.3
  • PCL >= 1.9
  • Boost >= 1.54
  • ROS >= Noetic
wget -O gtsam.zip https://github.com/borglab/gtsam/archive/refs/tags/4.1.1.zip
unzip gtsam.zip
cd gtsam-4.1.1/
mkdir build && cd build
cmake -DGTSAM_BUILD_WITH_MARCH_NATIVE=OFF -DGTSAM_USE_SYSTEM_EIGEN=ON ..
sudo make install -j16

git clone https://github.com/MIT-SPARK/TEASER-plusplus.git
cd TEASER-plusplus && mkdir build && cd build
cmake .. -DENABLE_DIAGNOSTIC_PRINT=OFF
sudo make install -j16
sudo ldconfig

Build Modules

  • tbb (is used for faster Quatro)
sudo apt install libtbb-dev
  • Build Quatro and nano_gicp
sudo apt install cmake libeigen3-dev libboost-all-dev

cd ~/your_workspace/src/VIP_ROBOT/third_party
git clone git@github.com:url-kaist/Quatro.git
git clone git@github.com:engcang/nano_gicp.git
cd ~/your_workspace

# nano_gicp, quatro first
catkin build nano_gicp -DCMAKE_BUILD_TYPE=Release

# build quatro
catkin build quatro -DCMAKE_BUILD_TYPE=Release -DQUATRO_TBB=ON -DQUATRO_DEBUG=OFF
catkin build -DCMAKE_BUILD_TYPE=Release

You have cloned the fast_lio source code by cloning our Repository in ./fast_lio

You can just run this code and get the whole Point Cloud map.

# build fast_lio
catkin build fast-lio

# Run the launch file based on your device (./fast_lio/launch)
roslaunch fast_lio $your_device_launch$.launch

FAST-LIO point cloud map (pcl_viewer) -> HKUST(GZ) W1 5F Maker Space FAST-LIO point cloud map

You have cloned the fast_lio_sam_qn source code by cloning our Repository in ./fast_lio_sam_qn

You can just run this code and get the whole Point Cloud map.

# build fast_lio_sam_qn 
catkin build fast_lio_sam_qn 

# You should change the yaml file in the launch based on your device (./fast_lio_sam_qn/launch/map_vipbot.launch)
roslaunch fast_lio_sam_qn map_vipbot.launch

FAST-LIO_SAM point cloud map (CloudCompare) -> HKUST(GZ) Core teaching area FAST-LIO_SAM point cloud map

You have cloned the fast_lio_localization_qn source code by cloning our Repository in ./fast_lio_localization_qn

You can just run this code and get the position data

# build fast_lio_localization_qn 
catkin build fast_lio_localization_qn

Edit the ./fast_lio_lolization_qn/config/vipbot_config.yaml: Line4: saved_map: for your result.bag path

# You should change the yaml file in the launch based on your device (./fast_lio_localization_qn/launch/run_loc_qn.launch)
roslaunch fast_lio_localization_qn run_loc_qn.launch

Localization in the point cloud map FAST-LIO_SAM point cloud map

RUN on the our Moving Platform

  • open roscore , and ouster os1 in supervise.html
  • open a terminal and run:
roslaunch real_v1_config ranger_mini_v2.launch
  • open a terminal and run:
roslaunch real_v1_config load_move_base.launch

and then operate in Rviz, like pub a destination. ​

Tutorials:

Introduce the logical structure and function of the component package.

We put the whole modules together to start with a project

navigation_config

All the running launch files are located in the navigation_config folder, and for convenience during final execution, each car is treated as a separate software package. The contents are listed in the following table:

Name Content Function
config move_base and lidar-related .yaml parameter file move_base and lidar-related parameter loading
launch move_baseand lidar-related.launchparameter file Load the parameters .yaml file & start the node
map .png and .yaml files for different sites Provide map_serve to publish 2d maps
pcd pcd point cloud map from fast_lio/fast_lio_sam 2D.png（stored locally because it is too large）
bag fast_lio_sam generates a .bag file that records keyframe Provides key frame positioning
trajectory Trajectory generated by topic_transit/TopicGenTrajectory CSV file Loaded in gbx_manual and published to 'move_base'
urdf/meshes Robot urdf file/simulation stl file Description of robot composition information/load model in simulation
rviz 'rviz' configuration file Save the files required by Rivz and fix the visual configuration

gbx_manual

In order to allow users to interact with each node and build a unified running process

STOP,MOVE,WAIT,PULL_OVER,ARRIVE 5 states with FSM for managing the entire navigation cycle。

Provides the /gbx_manual/pub_trajectory service interface externally. During initialization, the trajectory in navigation_config/real_v1_config/trajectory is loaded. Upon receiving a request containing correct trajectory information, the trajectory point stored by the trajectory point will be released to /click_point.

E3_121 is our test room

rosservice call /gbx_manual/pub_trajectory "sender: 'user' path_name: 'E3_121'"

This sends the request E3_121 to gbx_manual

How to run

Use roslaunch instead of mon launch to load fast_xxx related launch files

Version 1

roslaunch real_v1_config load_manual.launch

Run the load_manual.launch file to do the following

  • Load global_costmap, local_costmap and other configuration files required by move_base and start the node
  • Load the configuration file of map and launch the map_server node to publish 2d map information on the /map topic Load the required configuration files for fast_lio and fast_lio_localization_qn and start the node
  • Load static tf_publisher and robot urdf and robot_state_publisher
  • Load the cloud_transit configuration file and start the node
  • Load the ranger_mini_v2 configuration file and start the node (Mini ranger 2)
  • Load the gbx_manual configuration file and start the node

In addition, these nodes can also be run separately, as shown in the following example:

  • Only run fast_lio
roslaunch real_v1_config ouster_os1_mapping.launch
  • Only run fast_sam_localization_qn
roslaunch real_v1_config ouster_os1_loc_qn.launch
  • Only run mini ranger 2 controller
roslaunch real_v1_config ranger_mini_v2.launch

Steps

1. Mapping

Runfast_lio and fast_lio_sam for keyframe point cloud map .bag and .pcd file

roslaunch real_v1_config ouster_os1_sam.launch

2. Localization

Run fast_lio and fast_sam_localization_qn，load .bag file

roslaunch real_v1_config ouster_os1_loc_qn.launch

3. Generate 2d maps from pcd files

roslaunch pcd2pgm run.launch
rosrun map_server map_saver -f map_name

2D map for planner 2D map for planne

4.Load 2D map and move_base

roslaunch real_v1_config load_move_base.launch

5.Load manual

roslaunch real_v1_config load_manual.launch

Rosbag

We will publish the rosbag soon.

Contributing

Yujie ZHOU, Siyang LIU, Kai ZHANG

License

Null

Contact

For questions or inquiries, please contact Yujie ZHOU.

About

No description, website, or topics provided.

Resources

Readme
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

No packages published

Languages

  • C++ 90.2%
  • CMake 4.7%
  • C 4.2%
  • Other 0.9%
0