Reeds Shepp Path Library

Forked from the repository of Guan-Horng Liu. Modified some places:

1. Modify the path_type function to get the length (>0 for forward, <0 for backward) and curvature (>0 for left, <0 for right) of each segment.

Overview

The library implements the analytic Reeds Shepp path between two SE2 configurations. Reeds Sheep path¹ is defined as the shortest traveling path of the Reeds-Shepp Car, a car that can go both forward and backward with a constrained turning radius. Reeds-Sheep path can be viewed as a generalized version of Dubins path. The primary C++ source code comes from Open Motion Planning Library(OMPL). And the API of Python version (primarily a Cython wrapper) is similar to the Dubins library from Andrew Walker. More descriptions can be referred to the OMPL and section 15.3.2 of the book Planning Algorithms.

Installation

PyPI

$ pip install reeds-shepp

from source

$ sudo python setup.py install

Demo

$ python demos/demo.py

API

• path_length(q0, q1, turning_radius)

  • return total length of Reed Shepps curve from q0 to q1 with specified turning radius

• path_sample(q0, q1, turning_radius, step_size)

  • return list of tuple (x,y,theta) uniformly sampled from corresponding Reed Shepps curve

• path_type(q0, q1, turning_radius)

  • return the tuple of path segment type and length for the Reed Sheps curve from q0 to q1 with specified turning radius

References

Footnotes

1. Reeds, J., & Shepp, L. (1990). Optimal paths for a car that goes both forwards and backwards. Pacific journal of mathematics, 145(2), 367-393. ↩