🗺️ GMapping-SLAM Mapping

1. Course Content

• Master the complete process of building environment occupancy grid maps using the Gmapping laser SLAM algorithm.

• Control the A1 Pro robot to move and explore unknown areas via keyboard, and view mapping results in real-time in RViz2.

• Master the method for saving occupancy grid maps (.yaml + .pgm) after mapping is complete.

2. Gmapping Algorithm Introduction

• Basic Principle: Gmapping is an open-source laser SLAM algorithm based on the RBPF (Rao-Blackwellized Particle Filter) particle filter framework, using a localization and mapping separation mechanism (localize first, then update map).

• Main Improvements: Introduces improved proposal distribution and adaptive selective resampling, significantly reducing the required number of particles.

• Applicable Scenarios and Characteristics:

  • Advantages: Low computational overhead, strong real-time performance, and high accuracy when building small to medium-sized indoor environment maps.

  • Notes: Recommended single-frame laser points should be less than 1440; no global loop closure optimization mechanism, building ultra-large scenes requires increasing particle count, with memory and computing power consumption increasing accordingly.

3. Preparation and Running

This example requires opening 3~4 terminals to start the following nodes in sequence:

3.1 Start Gmapping Mapping Node

Run in Terminal 1 (automatically includes chassis motor driver, lidar driver, and Gmapping algorithm node):

xxxxxxxxxx
ros2 launch yahboomcar_nav map_gmapping_launch.py

3.2 Start RViz Map Visualization

Run in Terminal 2 (it is recommended to open RViz on the virtual machine if the VM and robot are on the same LAN):

xxxxxxxxxx
ros2 launch yahboomcar_nav display_map_launch.py

3.3 Start Keyboard Control to Move Robot for Mapping

Run in Terminal 3 the keyboard control node:

xxxxxxxxxx
ros2 run yahboomcar_ctrl yahboom_keyboard

3.4 Mapping Operation and Save Instructions

1. Remote Control Mapping Tips:

   • Focus on the keyboard control terminal, use keyboard keys to smoothly remote control the robot (you can press z / x to appropriately reduce linear and angular velocity).

   • Turn slowly at corners and corridor edges, try to bring the robot back to the starting point to form a physical loop, eliminating accumulated drift.

2. Save Occupancy Grid Map: When the scanned area is complete and the outline is clear, run the map save command in Terminal 4:

   xxxxxxxxxx
   ros2 launch yahboomcar_nav save_map_launch.py map_name:=map
   

   

   

   Map files will be saved in the following directory by default (generating map.yaml and map.pgm):

   xxxxxxxxxx
   ~/yahboomcar_ros2_ws/yahboomcar_ws/src/yahboomcar_nav/maps/
   

4. Core Source Code and Algorithm Parameter Analysis

4.1 Mapping Launch File Analysis

Launch File Path:

xxxxxxxxxx
$HOME/yahboomcar_ros2_ws/yahboomcar_ws/src/yahboomcar_nav/launch/slam/map_gmapping_launch.py

​x
def generate_launch_description():
    # Introduce chassis driver and lidar point cloud publishing
    laser_bringup_launch = IncludeLaunchDescription(PythonLaunchDescriptionSource(
        [os.path.join(get_package_share_directory('yahboomcar_nav'), 'launch','robot'),
         '/laser_bringup_launch.py'])
    )

    # Introduce Gmapping SLAM mapping node
    slam_gmapping_launch = IncludeLaunchDescription(PythonLaunchDescriptionSource(
        [os.path.join(get_package_share_directory('slam_gmapping'), 'launch'),
         '/slam_gmapping.launch.py'])
    )

    return LaunchDescription([laser_bringup_launch, slam_gmapping_launch])

• laser_bringup_launch.py: One-click start of A1 Pro chassis motor driver, IMU attitude solution, lidar ranging point cloud, and odometry TF transformation.

• slam_gmapping.launch.py: Load Gmapping algorithm core node and load configuration parameters.

4.2 Map Save Launch File Analysis

Launch File Path:

xxxxxxxxxx
$HOME/yahboomcar_ros2_ws/yahboomcar_ws/src/yahboomcar_nav/launch/slam/save_map_launch.py

This Launch node uniformly handles map saving:

• Default save directory is set to map_folder_dir = "~/yahboomcar_ros2_ws/yahboomcar_ws/src/yahboomcar_nav/maps".

• Calls nav2_map_server's map_saver_cli node, subscribes to /map topic and outputs occupancy grid matrix as standard PGM image and description file YAML:

xxxxxxxxxx
    save_map = Node(
        package='nav2_map_server',
        executable='map_saver_cli',
        parameters=[{
            'map_subscribe_transient_local': False,
            'save_map_timeout': 10.0,
        }],
        arguments=[
            '-f', PathJoinSubstitution([map_save_dir, map_name]),
            '--free', '0.196',
            '--occ', '0.65'],
        condition=IfCondition(PythonExpression(["'", map_type, "' == 'gridmap'"]))
    )

4.3 Core Algorithm and Parameter Interpretation

Parameter Configuration File Path:

xxxxxxxxxx
$HOME/yahboomcar_ros2_ws/software/library_ws/src/ros2_gmapping/slam_gmapping/params/slam_gmapping.yaml

Core Parameter Description Table

Parameter File Core Code Snippet

/slam_gmapping:
  ros__parameters:
    angularUpdate: 0.5            # Rotation update threshold (radians)
    linearUpdate: 1.0             # Linear displacement update threshold (meters)
    temporalUpdate: 1.0           # Periodic time update (seconds)
    particles: 50                 # Particle filter particle count
    delta: 0.05                   # Map grid resolution (meters/pixel)
    map_update_interval: 5.0      # Map update publishing period (seconds)
    maxUrange: 5.0                # Maximum effective laser distance for mapping (meters)
    maxRange: 8.0                 # Laser maximum truncation distance (meters)
    minimum_score: 0.1            # Scan matching minimum confidence score
    occ_thresh: 0.25              # Grid occupancy probability threshold
    resampleThreshold: 0.5        # Particle resampling threshold
    base_frame: base_footprint    # Robot base coordinate frame
    map_frame: map                # Global map coordinate frame
    odom_frame: odom              # Odometry coordinate frame