This repository contains a ROS package which localizes a robot in a known map with the use of a 2D LiDAR. Using the estimated pose and the ROS navigation stack, the robot is able to navigate through the environment.
An installation of the Robot Operating System (ROS) is required to run this package. This package has only been tested on the Kinetic and Melodic distributions of ROS.
Clone this repository into a catkin workspace src directory, and build the package.
$ git clone https://github.com/kenkangg/ROS_Monte_Carlo_Localization.git
$
$ cd ~/catkin_ws
$ catkin_make
$ source devel/setup.bash
The actual usage of this package consists of four steps: launching the Gazebo simulation, localizing the robot, and navigating it to a goal pose.
By launching the udacity_world.launch file, a robot, described in the urdf directory of this repository is placed in the jackal race environment. While this robot is equipped with both a planar laser rangefinder and an rgb camera, only the laser rangefinder is used for localization.
$ roslaunch udacity_bot udacity_world.launch
The amcl launch file starts the localization node which subscribes to the laser rangefinder, initial pose, robot transforms, and the predefined map. The output of this node is a particle cloud, estimated pose, and odometry transform information. By opening the rviz config file, this information can be easily visualized.
$ roslaunch udacity_bot amcl.launch
$ rosrun rviz rviz -d ~/catkin_ws/src/udacity_bot/udacity_bot.rviz
This file is not necessary to run localization, but will send the robot to the pose that is pictured at the beginning of this documentation. A custom goal pose can be chosen by clicking the 2D Nav Goal option at the top of the Rviz GUI and marking the desired location on the map.
$ rosrun udacity_bot navigation_goal