final_project

This repository is the implementation for ME5413 Autonomous Mobile Robotics Final Project

Created by Group 12

Dependencies

The requirement can be found in https://github.com/NUS-Advanced-Robotics-Centre/ME5413_Final_Project and also an extra ros package

• pcl_ros is used for generating pcl file from rostopic

Installation

This repo is a ros workspace, containing several rospkgs:

• ALOAM is SLAM algorithm package used for 3D mapping
• costmap_prohibition_layer provides a plugin to add virtual obstacle in costmap to set prohibited region
• interactive_tools are customized tools to interact with gazebo and your robot
• jackal_description contains the modified jackal robot model descriptions
• jackal_navigation contains a probabilistic localization system for navigation
• me5413_world the main pkg containing the gazebo world, and the launch files
• pcd2pgm for converting pcd file to pgm map which can be used for navigation

# Clone this repo
cd
git clone https://github.com/Tomshine123/My_project.git
cd My_project

# Install all dependencies
rosdep install --from-paths src --ignore-src -r -y

# Build
catkin_make
# Source 
source devel/setup.bash

To properly load the gazebo world, you will need to have the necessary model files in the ~/.gazebo/models/ directory, which installation instructions can be found in URL above.

Usage

1. Mapping

Following commands will launch the gazebo with the project world

# Launch Gazebo World together with our robot
roslaunch me5413_world world.launch

In the second terminal, launch A-LOAM SLAM for 3D Lidar

source devel/setup.bash
roslaunch aloam_velodyne aloam_velodyne_VLP_16.launch 

In order to drive the robot around for mapping, launch keyboard control node in the third terminal

rosrun teleop_twist_keyboard teleop_twist_keyboard.py 

By above you can visualize pointclouds, and finish mapping by exploring all possible areas To save the pointclouds in a pcl file, run rosnode in pcl_ros package to generate file in current direction

rosrun pcl_ros pointcloud_to_pcd input:=/laser_cloud_surround

To convert the pcl file to pgm format, run rosnode in pcd2pgm package after modify the path in test.cpp

private_nh.param("file_directory", file_directory, std::string("/home/tommy/pcd/")); # the path to pcd file

private_nh.param("file_name", file_name, std::string("12673.944000000")); # the name of the pcd file

Then build the workspace again and run the rosnode

cd ~/My_project
catkin_make
rosrun pcd2pgm pcd2topic

And run map saver in another terminal

rosrun map_server map_saver

After transforming you can obtain the global map by 3D Lidar SLAM like below and also a corresponding .yaml file， rename both as my_map and put them into src/me5413_world/maps/

2. Navigation

To begin with adding prohibited region in the map, find prohibition_areas.yaml in My_project/src/jackal_navigation/params/map_nav_params/ and modify the following content in the file to the desired one

prohibition_areas:

# define an obstacle point
# - [17.09, -6.388]
 
# define  an obstacle line
 - [[4.5, 1.5],
    [4.5, 5.0]]
 - [[1.0, 1.0],
    [3.0, 1.0]]

# define  an obstacle region
#- [[-11.15, -15.614],
#   [-12.35, -13.89],
#   [-10.05, -12.218]]

After that, also check if following lines in move_base.launch in jackal_navigation package,

<rosparam file="$(find jackal_navigation)/params/map_nav_params/prohibition_areas.yaml" command="load" ns="global_costmap/costmap_prohibition_layer" />
<rosparam file="$(find jackal_navigation)/params/map_nav_params/prohibition_areas.yaml" command="load" ns="local_costmap/costmap_prohibition_layer" />

Close all terminals and start new one in the workspace

source develop/setup.bash
roslaunch me5413_world world.launch

In the second terminal launch the navigation node

source develop/setup.bash
roslaunch me5413_world navigation.launch

Click the buttons of the bottom right panel in rviz and navigation to the specified pose

Get Support

If you have any questions please feel free to open a pull-request

License

The My_Project is released under the MIT License