deltarobustness's Introduction

deltaRobustness

This repository includes the code to algorithmically check $\delta$-robustness, as well as its application towards validating $\delta$-robustness for a bipedal walking gait.

Setup

The periodic orbits were generated using a modification of the FROST Toolbox (Modified Toolbox, Original Toolbox). Therefore, the modified toolbox is imcluded in the repository since it is needed to run the simulation code.

git clone --recursive https://github.com/maegant/deltaRobustness.git

Frost Setup

To setup FROST, please follow the installation instructions on the FROST website (https://ayonga.github.io/frost-dev/pages/installation.html). This includes adding the following two lines to your ~/.bashrc (make sure to update the path with the version of Mathematica you have)

LD_LIBRARY_PATH=/usr/local/Wolfram/Mathematica/12.0/SystemFiles/Links/MathLink/DeveloperKit/Linux-x86-64/CompilerAdditions:$LD_LIBRARY_PATH
LD_LIBRARY_PATH=/usr/local/Wolfram/Mathematica/12.0/SystemFiles/Links/WSTP/DeveloperKit/Linux-x86-64/CompilerAdditions:$LD_LIBRARY_PATH
export LD_LIBRARY_PATH

You can test if you've properly setup FROST by running the following initialization function in MATLAB

>> frost_addpath();

Main Script

To run the code, use the main_script.m script. I suggest running each section individually since some sections take several minutes to run and the figures may overwrite

Main Algorithm

The main implementation of the $\delta$-robustness optimization is contained within the @TestRobustness class in the function RunAlgorithm. Explicitly, the algorithm is as follows:

Recommend Projects

maegant / deltarobustness Goto Github PK