Skip to content
Multi-contact walking pattern generation over rough terrains by linearized MPC
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
cones
pymanoid @ 81874df
sep
staircase
stances
wpg
.gitignore
.gitmodules
LICENSE
README.md
walk.sh

README.md

Multi-contact linearized model predictive control

Source code for https://hal.archives-ouvertes.fr/hal-01349880/document

Installation

On Ubuntu 14.04, once you have installed OpenRAVE, do:

sudo apt-get install cython python python-dev python-pip python-scipy python-shapely
sudo pip install pycddlib quadprog pyclipper

Then, clone the repository and its submodule via:

git clone --recursive https://github.com/stephane-caron/3d-walking-lmpc.git

If you already have pymanoid installed on your system, be sure that its version matches that of the submodule.

Usage

The walk.sh script from the top-level directory opens the staircase simulation.

For a more detailed analysis, there are three subfolders in this repository, corresponding to different Sections of the paper:

  • Section IV: cones/ contains scripts to display and play with the CoM acceleration cones
  • Section VI: staircase/ walks the humanoid model around a circular staircase with tilted stepping stones
  • Appendix: sep/ compares four algorithms calculating the static-equilibrium polygon

Due to the copyright problem, we cannot release the COLLADA model HRP4R.dae used to produce the accompanying video and paper illustrations. We have replaced it with JVRC-1, which has the same kinematic chain.

You can’t perform that action at this time.