/mobile_final_group12

Primary LanguageC++MIT LicenseMIT

mobile_final_group12

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

Created by Group 12

Ubuntu 20.04 ROS Noetic C++ Python

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
  • teb_local_planner is used for local path plan, you need use the following command to install this package:
sudo apt-get install ros-noetic-teb-local-planner

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/zzf-zzf/mobile_final_group12.git
cd mobile_final_group12


# 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 pointcloud 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

pc_map 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/($ your_username)/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 ~/mobile_final_group12
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/

map

2. Navigation

To begin with adding prohibited region in the map, find prohibition_areas.yaml in mobile_final_group12/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" />

Then, add follow code to global_costmap_params.yaml:

plugins:
- {name: static_layer, type: "costmap_2d::StaticLayer"}
# - {name: obstacles_layer, type: "costmap_2d::ObstacleLayer"}
- {name: inflater_layer, type: "costmap_2d::InflationLayer"}
- {name: costmap_prohibition_layer, type: "costmap_prohibition_layer_namespace::CostmapProhibitionLayer"}

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 navigation

costmap

3. Random start and end point

If you want to change the initial position of robot and set different goal pose, you should change follow code at amcl.launch which is define the load position ar rviz:

<param name="initial_pose_x" value="($ x position you want)" />
<param name="initial_pose_y" value="($ y position you want)" />
<param name="initial_pose_a" value="($ angle you want)" />

Then change corresponding code at spawn_jackal.launch which define the position at gazebo:

<node name="urdf_spawner" pkg="gazebo_ros" type="spawn_model"
	      args="-urdf -model jackal -param robot_description -x 0 -y 0 -z 1 -R 0 -P 0 -Y 0" />
<!-- the value of parameter should same as `amcl.launch` -->

Random start

4. Change the navigation algorithm

If you want to use other gloabl or local planner algorithm, you should adjust the following code at move_base.launch:

<!-- <rosparam file="$(find jackal_navigation)/params/base_local_planner_params.yaml" command="load" /> -->
<!-- <rosparam file="$(find jackal_navigation)/params/dwa_local_planner_params.yaml" command="load" /> -->
<rosparam file="$(find jackal_navigation)/params/teb_local_planner_params.yaml" command="load" />

<!-- <param name="base_local_planner" value="base_local_planner/TrajectoryPlannerROS"/> -->
<!-- <param name="base_local_planner" value="dwa_local_planner/DWAPlannerROS" /> -->
<param name="base_local_planner" value="teb_local_planner/TebLocalPlannerROS" />

Our repo gives three local plan algorithm to use, the default is teb_local_planner, if you want to use other, you just need to comment now algorithm and select which you want, but if you want to use external algorithm, you not only need to change this parameters, but also need to add parameter file like teb_local_planner_params.yaml.

For global algorithm, the default is good, but if you prefer other algorithm, you can change following code at move_base.launch and add configuration file：

<param name="base_global_planner" type="string" value="navfn/NavfnROS" />

Get Support

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

License

The mobile_final_group12 is released under the MIT License