Skip to content

JasonChmn/multicontact-locomotion-planning

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

236 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This project implement the framework for mulicontact locomotion planning proposed in the loco-3d project. This framework decompose the global locomotion problem in several subproblems solved sequentially :

  • The contact planning
  • Centroidal trajectory optimization
  • Wholebody motion generation

It contains wrapper for the different blocks of the framework and script to automatically formulate problems, retrieve and store solutions and connect every blocks together. Along with a lot of helpers for visualization or exports.

The goal of this framework is to be modular, allowing the user to select a method for each of the subproblem or add a wrapper for any new method from the state-of-the-art solving one of this subproblem and connect it seamlessly to the rest of the framework.

Installation procedure :

This package rely on a lot of optionnal packages, see the section 'Available methods' for more information on the installation of this optional packages.

Install depencies from binary repositories :

  1. Add robotpkg to your apt configuration: http://robotpkg.openrobots.org/debian.html
  2. sudo apt update && sudo apt install robotpkg-py27-pinocchio robotpkg-py27-multicontact-api

Install depencies from sources :

Follow the instruction from https://github.com/loco-3d/multicontact-api

Installation procedure :

Once the depencies are correctly installed, clone the repository :

git clone https://github.com/loco-3d/multicontact-locomotion-planning.git

And add the directory multicontact-locomotion-planning/script to your PYTHONPATH.

Usage

TODO

Available methods :

Currently supported method for each subproblem, you need to install the packages linked separetely if you want to use this methods.

Contact generation :

Centroidal trajectory optimization :

Wholebody motion generation :

Effector trajectories :

  • Basic predefined splines
  • Limb-RRT (geometric path found by path-planning). Require RBPRM
  • Optimized limb-rrt (Dynamically consistent spline based on the results of limb-rrt)

Visualization :

Export :

About

Generation of whole body motion with the decoupled framework proposed in the Loco-3D project

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors

Languages

  • Python 96.5%
  • Shell 2.6%
  • Dockerfile 0.9%