小白也能懂!Python+AI加持,怎么用ROS2让机器人认路?

echo "source /opt/ros/humble/setup.bash" >> ~/.bashrcsource ~/.bashrc

# 安装Gazebo与ROS2接口包sudo apt install ros-humble-gazebo-ros-pkgs ros-humble-gazebo-ros2-control# 安装TurtleBot3相关包sudo apt install ros-humble-turtlebot3* ros-humble-teleop-twist-keyboard

# 安装SLAM Toolboxsudo apt install ros-humble-slam-toolbox# 安装Nav2导航栈及其启动文件sudo apt install ros-humble-navigation2 ros-humble-nav2-bringup# 安装其他实用工具（用于后续调试）sudo apt install ros-humble-tf2-tools ros-humble-rqt-tf-tree

ros2 pkg list | grep slam_toolboxros2 pkg list | grep nav2_bringup

mkdir -p ~/ros2_ws/srccd ~/ros2_wscolcon build --symlink-installecho "source ~/ros2_ws/install/setup.bash" >> ~/.bashrcsource ~/.bashrc

export TURTLEBOT3_MODEL=waffleros2 launch turtlebot3_gazebo turtlebot3_world.launch.py

# 新终端export TURTLEBOT3_MODEL=waffleros2 run teleop_twist_keyboard teleop_twist_keyboard

<?xml version="1.0"?><robotxmlns:xacro="http://www.ros.org/wiki/xacro"name="my_robot">    <!-- 定义颜色、尺寸等常量 -->    <xacro:propertyname="base_length"value="0.3" />    <xacro:propertyname="base_radius"value="0.1" />    <!-- 底盘 link -->    <linkname="base_link">        <visual>            <geometry><cylinderlength="${base_length}"radius="${base_radius}"/></geometry>            <materialname="blue"/>        </visual>        <collision>            <geometry><cylinderlength="${base_length}"radius="${base_radius}"/></geometry>        </collision>        <inertial>            <massvalue="2.0"/>            <inertiaixx="0.01"ixy="0.0"ixz="0.0"iyy="0.01"iyz="0.0"izz="0.01"/>        </inertial>    </link>    <!-- 左轮关节 -->    <jointname="left_wheel_joint"type="continuous">        <parentlink="base_link"/>        <childlink="left_wheel"/>        <originxyz="0 ${base_radius+wheel_width/2} 0"rpy="-1.5708 0 0"/>        <axisxyz="0 0 1"/>    </joint>    <!-- Gazebo 差分驱动插件 -->    <gazebo>        <pluginname="gazebo_ros_diff_drive"filename="libgazebo_ros_diff_drive.so">            <ros><namespace>/</namespace></ros>            <update_rate>50</update_rate>            <left_joint>left_wheel_joint</left_joint>            <right_joint>right_wheel_joint</right_joint>            <wheel_separation>${base_radius*2 + wheel_width}</wheel_separation>            <wheel_diameter>${wheel_radius*2}</wheel_diameter>            <command_topic>cmd_vel</command_topic>            <odometry_topic>odom</odometry_topic>            <odometry_frame>odom</odometry_frame>            <robot_base_frame>base_footprint</robot_base_frame>        </plugin>    </gazebo></robot>

ros2 run tf2_tools view_frames   # 生成frames.pdf

my_robot_description/├── CMakeLists.txt├── package.xml├── urdf/│   ├── my_robot.urdf.xacro│   └── materials.xacro├── meshes/└── launch/    ├── display.launch.py    └── spawn_robot.launch.py

import osfrom launch import LaunchDescriptionfrom launch_ros.actions import Nodefrom xacro import process_filedef generate_launch_description():    pkg_share = os.path.join(get_package_share_directory('my_robot_description'))    urdf_path = os.path.join(pkg_share, 'urdf', 'my_robot.urdf.xacro')    robot_description = process_file(urdf_path).toxml()    return LaunchDescription([        Node(package='robot_state_publisher', executable='robot_state_publisher',             parameters=[{'robot_description': robot_description}]),        Node(package='joint_state_publisher_gui', executable='joint_state_publisher_gui'),        Node(package='rviz2', executable='rviz2'),    ])

# mapper_params_online_async.yamlslam_toolbox:  ros__parameters:    odom_frame: odom    map_frame: map    base_frame: base_footprint    scan_topic: /scan    mode: mapping    minimum_range: 0.2    maximum_range: 10.0    minimum_travel_distance: 0.1    minimum_travel_heading: 0.2    do_loop_closing: true    loop_search_space: 8.0    map_update_interval: 5.0    enable_interactive_mode: true    # ... 其他参数

export TURTLEBOT3_MODEL=waffleros2 launch turtlebot3_gazebo turtlebot3_world.launch.py

ros2 launch slam_toolbox online_async_launch.py \  slam_params_file:=./src/my_robot_navigation/config/mapper_params_online_async.yaml \  use_sim_time:=true

ros2 run teleop_twist_keyboard teleop_twist_keyboard

ros2 run nav2_map_server map_saver_cli -f ~/maps/my_mapros2 service call /slam_toolbox/serialize_map slam_toolbox/srv/SerializePoseGraph \  "{filename: '/home/your_user/maps/my_pose_graph'}"

bt_navigator:  ros__parameters:    default_nav_to_pose_bt_xml: /opt/ros/humble/share/nav2_bt_navigator/behavior_trees/navigate_to_pose_w_replanning.xmlcontroller_server:  ros__parameters:    controller_frequency: 20.0    FollowPath:      plugin: "dwb_core::DWBLocalPlanner"      max_vel_x: 0.22      max_vel_theta: 1.0      path_distance_bias: 32.0      goal_distance_bias: 24.0local_costmap:  local_costmap:    ros__parameters:      global_frame: odom      rolling_window: true      width: 3      plugins: ["voxel_layer", "inflation_layer"]global_costmap:  global_costmap:    ros__parameters:      global_frame: map      plugins: ["static_layer", "obstacle_layer", "inflation_layer"]amcl:  ros__parameters:    global_frame_id: map    odom_frame_id: odom    laser_model_type: likelihood_field    min_particles: 500    max_particles: 2000

ros2 launch turtlebot3_gazebo turtlebot3_world.launch.py

ros2 launch nav2_bringup bringup_launch.py \  use_sim_time:=true \  map:=/home/your_user/maps/my_map.yaml \  params_file:=./src/my_robot_navigation/config/nav2_params.yaml

rviz2 -d /opt/ros/humble/share/nav2_bringup/rviz/nav2_default_view.rviz

mode: localizationmap_file_name: "/home/your_user/maps/my_pose_graph"

ros2 launch slam_toolbox online_async_launch.py \  slam_params_file:=./config/mapper_params_localization.yaml use_sim_time:=true

ros2 launch nav2_bringup navigation_launch.py use_sim_time:=true params_file:=./config/nav2_params.yaml