工作小笔记——安装turtlebot3的ROS gazebo仿真环境并使用cartographer进行建图和定位

---

前言

本文记录在之前已经安装完毕cartographer的workspace的基础上安装turtlebot3完整开发仿真环境的过程。

1. 环境要求

已经按照https://blog.csdn.net/m0_71775106/article/details/127707299文章的说明安装完成了相应的ros环境配置和cartographer。

2. 安装

cd ~/carto_ws/src
git clone https://github.com/ROBOTIS-GIT/turtlebot3.git
git clone https://github.com/ROBOTIS-GIT/turtlebot3_simulations.git
cd ..
rosdep install --from-paths src --ignore-src --rosdistro=${ROS_DISTRO} -y
catkin_make_isolated

echo “export SVGA_VGPU10=0” >> ~/.bashrc
source ~/.bashrc

3. 运行

3.1 启动仿真环境

新建一个Terminal窗口，输入如下命令：

source ~/carto_ws/devel_isolated/setup.bash
export TURTLEBOT3_MODEL=burger
roslaunch turtlebot3_gazebo turtlebot3_state_4.launch

3.2 建图

3.2.1 启动cartographer节点

新建一个Terminal窗口，运行如下命令

source ~/carto_ws/devel_isolated/setup.bash
export TURTLEBOT3_MODEL=burger
roslaucn turtlebot3_slam turtlebot3_slam.launch slam_methods:=cartographer configuration_basename:=turtlebot3_lds_2d_gazebo.lua

在使用gazebo仿真环境的时候，算法配置文件需要使用turtlebot3_lds_2d_gazebo.lua，而非turtlebot3_lds_2d.lua文件，两者区别如下：

22c22
<   tracking_frame = "imu_link", -- imu_link, If you are using gazebo, use 'base_footprint' (libgazebo_ros_imu's bug)
---
>   tracking_frame = "base_footprint", -- imu_link, If you are using gazebo, use 'base_footprint' (libgazebo_ros_imu's bug)
39c39
<   odometry_sampling_ratio = 0.1,
---
>   odometry_sampling_ratio = 1.,
41c41
<   imu_sampling_ratio = 0.1,
---
>   imu_sampling_ratio = 1.,

该launch文件还会启动rviz窗口，供调试查看。

3.2.2 移动机器人进行建图

新建一个Terminal窗口，运行如下命令打开键盘控制界面，可以通过键盘控制机器人进行移动，从而完成建图。

source ~/carto_ws/devel_isolated/setup.bash
export TURTLEBOT3_MODEL=burger
roslaunch turtlebot3_teleop turtlebot3_teleop_key.launch

3.2.3 保存地图

当地图建立完整后，新建一个Terminal串口，运行如下命令，保存地图：

source ~/carto_ws/devel_isolated/setup.bash
export TURTLEBOT3_MODEL=burger
# Finish the trajectory before saving the map
rosservice call /finish_trajectory 0
# Save the pbstream to a local file
rosservice call /write_state "{filename: '/home/<username>/turtlebot3_cartographer_stage_4.pbstream', include_unfinished_submaps: true}"
# Convert the pbstream format to ros map format (pgm  and yaml)
rosrun cartographer_ros cartographer_pbstream_to_ros_map -map_filestem=/home/<username>/turtlebot3_cartographer_stage_4 -pbstream_filename=/home/<username>/turtlebot3_cartographer_stage_4.pbstream -resolution=0.05

3.3 导航定位

3.3.1 准备文件

新建文件，~/carto_ws/src/turtlebot3/turtlebot3_slam/config/turtlebot3_lds_2d_gazebo_localization.lua，输入如下内容：

include "turtlebot3_lds_2d_gazebo.lua"

--TRAJECTORY_BUILDER.pure_localization = true
TRAJECTORY_BUILDER.pure_localization_trimmer = {
  max_submaps_to_keep = 3,
}
POSE_GRAPH.optimize_every_n_nodes = 50

return options

新建文件，~/carto_ws/src/turtlebot3/turtlebot3_slam/launch/turtlebot3_cartographer_localization.launch，输入如下内容（将其中<username>改为实际值）：

<launch>
  <!-- Arguments -->
  <arg name="model" default="$(env TURTLEBOT3_MODEL)" doc="model type [burger, waffle, waffle_pi]"/>
  <arg name="configuration_basename" default="turtlebot3_lds_2d_gazebo_localization.lua"/>
  <arg name="open_rviz" default="true"/>
  <arg name="load_state_filename" default="/home/<username>/turtlebot3_cartographer_stage_4.pbstream"/>

  <!-- TurtleBot3 -->
  <include file="$(find turtlebot3_bringup)/launch/turtlebot3_remote.launch">
    <arg name="model" value="$(arg model)" />
  </include>

  <!-- move_base -->
  <include file="$(find turtlebot3_navigation)/launch/move_base.launch">
    <arg name="model" value="$(arg model)" />
  </include>

  <!-- cartographer_node -->
  <node pkg="cartographer_ros" type="cartographer_node" name="cartographer_node" 
        args="-configuration_directory $(find turtlebot3_slam)/config
              -configuration_basename $(arg configuration_basename)
              -load_state_filename $(arg load_state_filename)"
        output="screen">
    <remap from="/imu" to="/flat_imu"/>
    <!--remap from="points2" to="/camera/depth/points" / -->
  </node>

  <!-- cartographer_occupancy_grid_node -->
  <node pkg="cartographer_ros" type="cartographer_occupancy_grid_node"
        name="cartographer_occupancy_grid_node" 
        args="-resolution 0.05" />

  <!-- flat_world_imu_node -->
  <node pkg="turtlebot3_slam" type="flat_world_imu_node" name="flat_world_imu_node" output="screen">
    <remap from="imu_in" to="/imu" />
    <remap from="imu_out" to="/flat_imu" />
  </node>

  <!-- rviz -->
  <group if="$(arg open_rviz)"> 
    <node pkg="rviz" type="rviz" name="rviz" required="true"
          args="-d $(find turtlebot3_slam)/rviz/turtlebot3_$(arg slam_methods).rviz"/>
  </group>
</launch>

3.3.2 运行导航节点

新建一个Terminal，输入如下命令：

source ~/carto_ws/devel_isolated/setup.bash
export TURTLEBOT3_MODEL=burger
roslaunch turtlebot3_slam turtlebot3_cartographer_localization.launch slam_methods:=cartographer

3.4 节点关系图

运行如下命令，可以打开节点关系图进行查看。

rqt_graph

参考文献

• 工作小笔记——基于ROS Melodic的cartographer安装, https://blog.csdn.net/m0_71775106/article/details/127707299

附录 其他几个包的简单安装说明

在目前的ros开发环境，可以轻松的安装其他算法包，这里简单列出几个。

附录A 安装husky机器人环境

A.1安装基础包

cd ~/carto_ws/src
git clone -b melodic-devel https://github.com/husky/husky.git
cd ..
rosdep install --from-paths src --ignore-src --rosdistro=${ROS_DISTRO} -y
catkin_make_isolated

echo “export SVGA_VGPU10=0” >> ~/.bashrc
source ~/.bashrc

A.2 安装cartographer支持

cd ~/carto_ws/src
git clone -b melodic-devel https://github.com/husky/husky-cartographer.git
cd ..
rosdep install --from-paths src --ignore-src --rosdistro=${ROS_DISTRO} -y
catkin_make_isolated

A.3 运行

source ~/carto_ws/devel_isolated/setup.bash
export HUSKY_URDF_EXTRAS=$(rospack find husky_description)/urdf/empty.urdf
roslaunch husky_gazebo husky_empty_world.launch
roslaunch teleop_twist_joy teleop.launch

附录B. 安装GVINS

B.1 安装

cd ~/carto_ws/src
git clone https://github.com/HKUST-Aerial-Robotics/gnss_comm.git
git clone https://github.com/HKUST-Aerial-Robotics/GVINS.git
cd ..
rosdep install --from-paths src --ignore-src --rosdistro=${ROS_DISTRO} -y
catkin_make_isolated

B.2 运行

source ~/carto_ws/devel_isolated/setup.bash
roslaunch gvins visensor_f9p.launch
rviz -d ~/carto_ws/src/GVINS/config/gvins_rviz_config.rviz
rosbag play sports_field.bag

上述命令中，rosbag回放文件可以从https://github.com/HKUST-Aerial-Robotics/GVINS-Dataset下载。