vgraph_nav

ROS package for a simple navigation system utilizing the Visibility Graph algorithm.
It is compatible with occupancy grid maps generated by popular 2D SLAM such as Cartographer, slam_toolbox, and slam_gmapping

Setup ⚙

$ cd {ROS_WORKSPACE}/src
$ git clone -b noetic-devel https://github.com/Alpaca-zip/vgraph_nav.git
$ wstool merge vgraph_nav/.rosinstall
$ wstool update
$ python3 -m pip install -r vgraph_nav/requirements.txt
$ rosdep install -r -y -i --from-paths .
$ catkin build

Run 🚀

Example of navigation using TurtleBot3.

$ roslaunch vgraph_nav turtlebot3_vgraph_navigation.launch

1. 2D Pose Estimate

Before starting navigation, you can define the starting point of the robot.
This is achieved using the 2D Pose Estimate button on the RViz.

More info : e-Manual

2. 2D Nav Goal

After setting the initial pose, the next step is to specify the target position using 2D Nav Goal.

More info : e-Manual

NOTE: The first time after launch, optimization using the Visibility Graph algorithm takes few minutes.

vgraph_planner_node 🤖

Params

• ~map_file: Path to the configuration file (yaml) for the occupancy grid map.
• ~test_folder: A debug folder primarily used for debugging the Visibility Graph algorithm.
• ~epsilon_factor: Adjust the corner detection accuracy of the occupancy grid map. Higher values reduce the number of corners.
• ~robot_radius: Specifies the radius of the robot. This is mainly used for obstacle detection.
• ~odom_topic: The name of the odometry topic.
• ~scan_topic: The name of the LaserScan topic.
• ~vel_linear: Linear velocity.
• ~vel_theta: Maximum angular velocity.
• ~angle_tolerance: During navigation, the robot will attempt to align its direction with the goal if the angle difference exceeds this tolerance.
• ~goal_tolerance: Distance to a waypoint at which it is considered reached.
• ~verbose: If True, the results of the optimization are displayed.

Topics

• Subscribed Topics:

  • /initialpose (geometry_msgs/PoseWithCovarianceStamped) : Receives the robot's initial pose from RViz's 2D Pose Estimate
  • /move_base_simple/goal (geometry_msgs/PoseStamped) : Receives navigation goals from RViz's 2D Nav Goal
  • /odom (nav_msgs/Odometry) : Provides the robot's odometry data. Can be remapped via ~odom_topic.
  • /scan (sensor_msgs/LaserScan) : Provides the 2D Lidar data of the robot. Can be remapped via ~scan_topic.

• Published Topics:

  • /cost_map (nav_msgs/OccupancyGrid) : Generates a path along the obstacles in this occupancy grid map. Mainly for visualization in RViz.
  • /path (nav_msgs/Path) : The path that the robot will follow. Mainly for visualization in RViz.
  • /cmd_vel (geometry_msgs/Twist) : Conveys the target velocity data of the robot.