ROS Motion Planning

Robot Motion planning is a computational problem that involves finding a sequence of valid configurations to move the robot from the source to the destination. Generally, it includes Path Searching and Trajectory Optimization.

• Path Searching: Based on path constraints, such as obstacles, to find the optimal sequence for the robot to travel from the source to the destination without any collisions.

• Trajectory Planning: Based on kinematics, dynamics and obstacles, it optimizes the trajectory of the motion state from the source to the destination according to the path.

This repository provides the implementation of common Motion Planning algorithms. The theory analysis can be found at motion-planning. Furthermore, we provide Python and MATLAB version.

Your stars, forks and PRs are welcome!

Contents

• Quick Start within 3 Minutes
• File Tree
• Tool Chains
• Version
• Papers
• Acknowledgments
• License
• Maintenance

0. Quick Start within 3 Minutes

Tested on ubuntu 20.04 LTS with ROS Noetic.

1. Install ROS (Desktop-Full suggested).

2. Install git.

   sudo apt install git

3. Other dependence.

   sudo apt install python-is-python3 \
   ros-noetic-amcl \
   ros-noetic-base-local-planner \
   ros-noetic-map-server \
   ros-noetic-move-base \
   ros-noetic-navfn

4. Clone the reposity.

   git clone https://github.com/ai-winter/ros_motion_planning.git

5. Compile the code.

   cd ros_motion_planning/
   catkin_make
   # or catkin build
   # you may need to install it by: sudo apt install python-catkin-tools

6. Execute the code.

   cd scripts/
   ./main.sh

   NOTE: Modifying certain launch files does not have any effect, as they are regenerated based on the src/user_config/user_config.yaml by a Python script when you execute main.sh. Therefore, you should modify configurations in user_config.yaml instead of launch files.

7. Use 2D Nav Goal in RViz to select the goal.

8. Moving!

9. You can use the other script to shutdown them rapidly.

   ./killpro.sh

10. For more details on system configuration, please refer to Configuration.

1. File Tree

The file structure is shown below.

ros_motion_planner
├── assets
├── scripts
└── src
    ├── core
    │   ├── global_planner
    │   ├── local_planner
    │   ├── curve_generation
    │   └── utils
    ├── sim_env             # simulation environment
    │   ├── config
    │   ├── launch
    │   ├── maps
    │   ├── meshes
    │   ├── models
    │   ├── rviz
    │   ├── urdf
    │   └── worlds
    ├── third_party
    │   ├── dynamic_rviz_config
    │   ├── dynamic_xml_config
    │   ├── gazebo_plugins
    │   └── rviz_plugins
    └── user_config         # user configure file

02. Tool Chains

For the efficient operation of the motion planning system, we provide a series of user-friendly simulation tools that allow for on-demand selection of these lightweight repositories.

Tool Version	Introduction
	This is a Gazebo plugin for pedestians with collision property. You can construct a dynamic environment in ROS easily using plugin.
	This repository provides a ROS-based visualization Rviz plugins for path planning and curve generation algorithms.

03. Version

Global Planner

Planner	Version	Animation
GBFS
Dijkstra
A*
JPS
D*
LPA*
D* Lite
Voronoi
Theta*
Lazy Theta*
Hybrid A*
RRT
RRT*
Informed RRT
RRT-Connect

Local Planner

Planner	Version	Animation
PID
DWA
APF
RPP
TEB
MPC
Lattice

Intelligent Algorithm

Planner	Version	Animation
ACO
GA
PSO
ABC

Curve Generation

Planner	Version	Animation
Polynomia
Bezier
Cubic Spline
BSpline
Dubins
Reeds-Shepp

04. Papers

Search-based Planning

• A*: A Formal Basis for the heuristic Determination of Minimum Cost Paths.
• JPS: Online Graph Pruning for Pathfinding On Grid Maps.
• Lifelong Planning A*: Lifelong Planning A*.
• D*: Optimal and Efficient Path Planning for Partially-Known Environments.
• D* Lite: D* Lite.
• Theta*: Theta*: Any-Angle Path Planning on Grids, Any-angle path planning on non-uniform costmaps.
• Lazy Theta*: Lazy Theta*: Any-Angle Path Planning and Path Length Analysis in 3D.
• Hybrid A*: Practical search techniques in path planning for autonomous driving

Sample-based Planning

• RRT: Rapidly-Exploring Random Trees: A New Tool for Path Planning.
• RRT-Connect: RRT-Connect: An Efficient Approach to Single-Query Path Planning.
• RRT*: Sampling-based algorithms for optimal motion planning.
• Informed RRT*: Optimal Sampling-based Path Planning Focused via Direct Sampling of an Admissible Ellipsoidal heuristic.

Evolutionary-based Planning

• ACO: Ant Colony Optimization: A New Meta-Heuristic.
• GA: Adaptation in Natural and Artificial Systems
• PSO: Particle Swarm Optimization

Local Planning

• DWA: The Dynamic Window Approach to Collision Avoidance.
• APF: Real-time obstacle avoidance for manipulators and mobile robots.
• RPP: Regulated Pure Pursuit for Robot Path Tracking

Curve Generation

• Reeds Shepp: Optimal paths for a car that goes both forwards and backwards
• Dubins: On curves of minimal length with a constraint on average curvature, and with prescribed initial and terminal positions and tangents

05. Acknowledgments

• Our robot and world models are from Dataset-of-Gazebo-Worlds-Models-and-Maps and aws-robomaker-small-warehouse-world. Thanks for these open source models sincerely.

• Pedestrians in this environment are using social force model(sfm), which comes from https://github.com/robotics-upo/lightsfm.

• A ROS costmap plugin for dynamicvoronoi presented by Boris Lau.

06. License

The source code is released under GPLv3 license.

07. Maintenance

Feel free to contact us if you have any question.