Description

A 3D collision-free path planning algorithm based on RRT and SP-RRT specialized for 8-link hyper-redundant robot implemented using ROS.

Distro: Ros neotic

The algorithm finds a 3D optimized path for obstacle environments. The visualtization is implemented in RVIZ and the majority code is written in C++. This package also offers implementation for using Octomap as obstacle-environment.

The package has four executables:

1. rrt_node
   
2. sp_rrt node
   
3. env_node
   
4. sp_rrt_node_octomap
   

RVIZ parameters:

1. Frame_id = "/path_planner"
   
2. Marker_topic = "path_planner_rrt"
   
3. MarkerArray_topic = "/occupied_cells_vis_array"
   

Instructions for plain RRT and SP-RRT:

1. Open terminal and type
   roscore
   
2. Open new terminal and go to the the root of your catkin workspace
   catkin_make
source ./devel/setup.bash
rosrun path_planner env_node
   
3. open new terminal
   rosrun rviz rviz
   
4. In the RVIZ window, change:
   fixed frame under global option to "path_planner"
   add a marker and change marker topic to "path_planner_rrt"
   
5. Open new terminal
   To run RRT algorithm: rosrun path_planner rrt_node
   To run SP_RRT algorithm: rosrun path_planner sp_rrt_node
   

Instructions for SP-RRT in Octomap:

Package Requirement: Octomap
Installation: sudo apt-get install ros-noetic-octomap*

1. Open file launch/octomap_tracking_server.launch and change <arg name="path" default="/XXX/src/path_planning/octomap/freiburg1_360.bt"/> into your own root.
   
2. Open terminal and type
   catkin_make
source ./devel/setup.bash
rosrun path_planner env_node
   
3. open new termianl
   roslaunch path_planning octomap_tracking_server.launch
   
4. open new terminal
   rosrun rviz rviz
   
5. open new terminal
   rosrun path_planning sp_rrt_octomap
   
6. In the RVIZ window, change:
   fixed frame under global option to "path_planner"
   add a Marker and change marker topic to "path_planner_rrt"
   add a MarkerArray and change topic to "occupied_cells_vis_array"

Reference:

Code: based on a planar RRT implementation from https://github.com/nalin1096/path_planning.
SP-RRT: H. Wei, Y. Zheng and G. Gu, "RRT-Based Path Planning for Follow-the-Leader Motion of Hyper-Redundant Manipulators," 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021, pp. 3198-3204.
Octomap: fr_078_tidyup.bt and freiburg1_360.bt are from Autonome Intelligente Systeme, Albert-Ludwigs-Universität Freiburg. The relevent link is http://ais.informatik.uni-freiburg.de/projects/datasets/octomap/.

Visualization Display

[Fig1. plain sp-rrt]

[Fig.2 sp-rrt in octomap]