Skip to content

giorgia-nadizar/VSREvoDevo

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

VSREvoDevo

This repository hosts the code used in the paper "On the Schedule for Morphological Development of Evolved Modular Soft Robots". The scope of the paper concerned the investigation of the effects of different developmental schedules on evolved Voxel-Based Soft Robots (VSRs), experimenting with various representations for development functions and robotic controllers.

Content

The content of this repository is organized into two main packages:

  • 2dhmsr, taken from 2D highly modular soft robots, where all the elements required to perform the simulation of VSRs are included; and
  • evolution, which contains all the components used to perform the evolutionary optimization, i.e., the evolution, of VSRs. This also includes a dependency to JGEA, a general evolutionary algorithm (EA) framework written in Java, used for actually performing the optimization part. The jar for JGEA is already included in the lib folder.

Suggested usage

Visualizing a VSR

To visualize a developing VSR you need to run the DevelopmentExample class, which is included in the 2dhmsr package. This will start a simulation and will display a video of a comb shaped VSR moving downhill and developing, i.e., gaining additional voxels, at regular time intervals. Note that in this case we are using a non-optimized VSR, hence most of its successful movement is caused by the inclination of the terrain.

Performing an optimization

Evolutionary optimizations are performed by running the Starter class contained in evolution.devolocomotion. This will start the evolution of a development function for VSRs with the default parameters (listed at the beginning of the method run()). One can freely play with the parameters, which are mostly self-explanatory, either changing the source code or by using the corresponding key words from command line. The most important ones are the devoFunction parameter, which describes the type of representation, and the evolver parameter, related to the EA used to perform the optimization. It is fundamental that these two parameters refer to the same genotype, e.g., both devoFunction and evolver should operate on a List<Double> genotype, otherwise an Exception is thrown.

Bibliography

Nadizar, Medvet, Miras. "On the Schedule for Morphological Development of Evolved Modular Soft Robots", European Conference on Genetic Programming (Part of EvoStar), 2022

@inproceedings{nadizar2022schedule,
  title={On the Schedule for Morphological Development of Evolved Modular Soft Robots},
  author={Nadizar, Giorgia and Medvet, Eric and Miras, Karine},
  booktitle={European Conference on Genetic Programming (Part of EvoStar)},
  pages={146--161},
  year={2022},
  organization={Springer}
}

About

This repository hosts the code used in the paper "On the Schedule for Morphological Development of Evolved Modular Soft Robots"

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages