Efficient neuroevolution for a quadruped robot

Shengbo Xu, Hirotaka Moriguchi, Shinichi Honiden

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

In this research, we investigate whether CoSyNE and CMA-NeuroES algorithms can efficiently optimize neural policy of a quadruped robot. Both of these algorithms are proven to optimize connection weights efficiently on Pole Balancing benchmark. Due to their good results on that benchmark, they are expected to be efficient on other control problems like gait generation. In this research we experimentally show that CMA-NeuroES have higher scalability to optimize Artificial Neural Networks for generating gaits of quadruped robots in comparison with CoSyNE. The results can be helpful for researchers and practitioners to choose the optimal Neuroevolution algorithm for generating gaits.

Original languageEnglish
Title of host publicationSimulated Evolution and Learning - 9th International Conference, SEAL 2012, Proceedings
Pages361-370
Number of pages10
DOIs
Publication statusPublished - 2012 Dec 26
Externally publishedYes
Event9th International Conference on Simulated Evolution and Learning, SEAL 2012 - Hanoi, Viet Nam
Duration: 2012 Dec 162012 Dec 19

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7673 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other9th International Conference on Simulated Evolution and Learning, SEAL 2012
CountryViet Nam
CityHanoi
Period12/12/1612/12/19

Fingerprint

Neuroevolution
Gait
Robot
Optimise
Robots
Benchmark
Balancing
Artificial Neural Network
Pole
Scalability
Poles
Control Problem
Choose
Neural networks

Keywords

  • CMA-ES
  • CMA-NeuroES
  • CoSyNE
  • evolution
  • neural network
  • neuroevolution
  • Simplex

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Xu, S., Moriguchi, H., & Honiden, S. (2012). Efficient neuroevolution for a quadruped robot. In Simulated Evolution and Learning - 9th International Conference, SEAL 2012, Proceedings (pp. 361-370). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7673 LNCS). https://doi.org/10.1007/978-3-642-34859-4_36

Efficient neuroevolution for a quadruped robot. / Xu, Shengbo; Moriguchi, Hirotaka; Honiden, Shinichi.

Simulated Evolution and Learning - 9th International Conference, SEAL 2012, Proceedings. 2012. p. 361-370 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7673 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Xu, S, Moriguchi, H & Honiden, S 2012, Efficient neuroevolution for a quadruped robot. in Simulated Evolution and Learning - 9th International Conference, SEAL 2012, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7673 LNCS, pp. 361-370, 9th International Conference on Simulated Evolution and Learning, SEAL 2012, Hanoi, Viet Nam, 12/12/16. https://doi.org/10.1007/978-3-642-34859-4_36
Xu S, Moriguchi H, Honiden S. Efficient neuroevolution for a quadruped robot. In Simulated Evolution and Learning - 9th International Conference, SEAL 2012, Proceedings. 2012. p. 361-370. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-34859-4_36
Xu, Shengbo ; Moriguchi, Hirotaka ; Honiden, Shinichi. / Efficient neuroevolution for a quadruped robot. Simulated Evolution and Learning - 9th International Conference, SEAL 2012, Proceedings. 2012. pp. 361-370 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{3b43709b7074497d88513146a3050d09,
title = "Efficient neuroevolution for a quadruped robot",
abstract = "In this research, we investigate whether CoSyNE and CMA-NeuroES algorithms can efficiently optimize neural policy of a quadruped robot. Both of these algorithms are proven to optimize connection weights efficiently on Pole Balancing benchmark. Due to their good results on that benchmark, they are expected to be efficient on other control problems like gait generation. In this research we experimentally show that CMA-NeuroES have higher scalability to optimize Artificial Neural Networks for generating gaits of quadruped robots in comparison with CoSyNE. The results can be helpful for researchers and practitioners to choose the optimal Neuroevolution algorithm for generating gaits.",
keywords = "CMA-ES, CMA-NeuroES, CoSyNE, evolution, neural network, neuroevolution, Simplex",
author = "Shengbo Xu and Hirotaka Moriguchi and Shinichi Honiden",
year = "2012",
month = "12",
day = "26",
doi = "10.1007/978-3-642-34859-4_36",
language = "English",
isbn = "9783642348587",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "361--370",
booktitle = "Simulated Evolution and Learning - 9th International Conference, SEAL 2012, Proceedings",

}

TY - GEN

T1 - Efficient neuroevolution for a quadruped robot

AU - Xu, Shengbo

AU - Moriguchi, Hirotaka

AU - Honiden, Shinichi

PY - 2012/12/26

Y1 - 2012/12/26

N2 - In this research, we investigate whether CoSyNE and CMA-NeuroES algorithms can efficiently optimize neural policy of a quadruped robot. Both of these algorithms are proven to optimize connection weights efficiently on Pole Balancing benchmark. Due to their good results on that benchmark, they are expected to be efficient on other control problems like gait generation. In this research we experimentally show that CMA-NeuroES have higher scalability to optimize Artificial Neural Networks for generating gaits of quadruped robots in comparison with CoSyNE. The results can be helpful for researchers and practitioners to choose the optimal Neuroevolution algorithm for generating gaits.

AB - In this research, we investigate whether CoSyNE and CMA-NeuroES algorithms can efficiently optimize neural policy of a quadruped robot. Both of these algorithms are proven to optimize connection weights efficiently on Pole Balancing benchmark. Due to their good results on that benchmark, they are expected to be efficient on other control problems like gait generation. In this research we experimentally show that CMA-NeuroES have higher scalability to optimize Artificial Neural Networks for generating gaits of quadruped robots in comparison with CoSyNE. The results can be helpful for researchers and practitioners to choose the optimal Neuroevolution algorithm for generating gaits.

KW - CMA-ES

KW - CMA-NeuroES

KW - CoSyNE

KW - evolution

KW - neural network

KW - neuroevolution

KW - Simplex

UR - http://www.scopus.com/inward/record.url?scp=84871382850&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871382850&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-34859-4_36

DO - 10.1007/978-3-642-34859-4_36

M3 - Conference contribution

SN - 9783642348587

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 361

EP - 370

BT - Simulated Evolution and Learning - 9th International Conference, SEAL 2012, Proceedings

ER -