A comparison of analytical and empirical controllers for the SLIP hopper

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

Abstract

The Spring-Loaded Inverted Pendulum (SLIP) model poses a challenging control problem, important for the development of legged robots, due to the difficultly in solving the stance phase of the dynamics. Multiple attempts have been made to approximate these dynamics to allow for an analytical control method; here four of these methods have been compared for controlling agile hopping, where there are large changes in forward velocity across a single stance. In addition, a new, empirical, approach has been demonstrated. In this, a simple control law is formulated, based on some simple approximations, which allows the parameters to be selected empirically through simulation. This has led to a controller able to offer similar performance to the best analytical approximation but with a much simpler form. This empirical controller may present new opportunities for controlling more complex dynamics and the development of a self-tuning method in future work.

LanguageEnglish
Title of host publicationTowards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings
Place of PublicationGermany
PublisherSpringer Verlag
Pages79-85
Number of pages7
ISBN (Print)9783319641065
DOIs
StatusE-pub ahead of print - 20 Jul 2017
Event18th Annual Conference on Towards Autonomous Robotic Systems, TAROS 2017 - Guildford, UK United Kingdom
Duration: 19 Jul 201721 Jul 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10454

Conference

Conference18th Annual Conference on Towards Autonomous Robotic Systems, TAROS 2017
CountryUK United Kingdom
CityGuildford
Period19/07/1721/07/17

Fingerprint

Hoppers
Inverted Pendulum
Pendulums
Controller
Controllers
Legged Robots
Self-tuning
Analytical Approximation
Complex Dynamics
Best Approximation
Control Problem
Tuning
Robots
Approximation
Simulation
Model

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Hale, M. F., du Bois, J. L., & Iravani, P. (2017). A comparison of analytical and empirical controllers for the SLIP hopper. In Towards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings (pp. 79-85). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10454 ). Germany: Springer Verlag. https://doi.org/10.1007/978-3-319-64107-2_7

A comparison of analytical and empirical controllers for the SLIP hopper. / Hale, Matthew F.; du Bois, Jonathan L.; Iravani, Pejman.

Towards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings. Germany : Springer Verlag, 2017. p. 79-85 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10454 ).

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

Hale, MF, du Bois, JL & Iravani, P 2017, A comparison of analytical and empirical controllers for the SLIP hopper. in Towards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10454 , Springer Verlag, Germany, pp. 79-85, 18th Annual Conference on Towards Autonomous Robotic Systems, TAROS 2017, Guildford, UK United Kingdom, 19/07/17. https://doi.org/10.1007/978-3-319-64107-2_7
Hale MF, du Bois JL, Iravani P. A comparison of analytical and empirical controllers for the SLIP hopper. In Towards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings. Germany: Springer Verlag. 2017. p. 79-85. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-64107-2_7
Hale, Matthew F. ; du Bois, Jonathan L. ; Iravani, Pejman. / A comparison of analytical and empirical controllers for the SLIP hopper. Towards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings. Germany : Springer Verlag, 2017. pp. 79-85 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{abefc77c877e484d975d556ae33511ab,
title = "A comparison of analytical and empirical controllers for the SLIP hopper",
abstract = "The Spring-Loaded Inverted Pendulum (SLIP) model poses a challenging control problem, important for the development of legged robots, due to the difficultly in solving the stance phase of the dynamics. Multiple attempts have been made to approximate these dynamics to allow for an analytical control method; here four of these methods have been compared for controlling agile hopping, where there are large changes in forward velocity across a single stance. In addition, a new, empirical, approach has been demonstrated. In this, a simple control law is formulated, based on some simple approximations, which allows the parameters to be selected empirically through simulation. This has led to a controller able to offer similar performance to the best analytical approximation but with a much simpler form. This empirical controller may present new opportunities for controlling more complex dynamics and the development of a self-tuning method in future work.",
author = "Hale, {Matthew F.} and {du Bois}, {Jonathan L.} and Pejman Iravani",
year = "2017",
month = "7",
day = "20",
doi = "10.1007/978-3-319-64107-2_7",
language = "English",
isbn = "9783319641065",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "79--85",
booktitle = "Towards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings",
address = "Germany",

}

TY - GEN

T1 - A comparison of analytical and empirical controllers for the SLIP hopper

AU - Hale, Matthew F.

AU - du Bois, Jonathan L.

AU - Iravani, Pejman

PY - 2017/7/20

Y1 - 2017/7/20

N2 - The Spring-Loaded Inverted Pendulum (SLIP) model poses a challenging control problem, important for the development of legged robots, due to the difficultly in solving the stance phase of the dynamics. Multiple attempts have been made to approximate these dynamics to allow for an analytical control method; here four of these methods have been compared for controlling agile hopping, where there are large changes in forward velocity across a single stance. In addition, a new, empirical, approach has been demonstrated. In this, a simple control law is formulated, based on some simple approximations, which allows the parameters to be selected empirically through simulation. This has led to a controller able to offer similar performance to the best analytical approximation but with a much simpler form. This empirical controller may present new opportunities for controlling more complex dynamics and the development of a self-tuning method in future work.

AB - The Spring-Loaded Inverted Pendulum (SLIP) model poses a challenging control problem, important for the development of legged robots, due to the difficultly in solving the stance phase of the dynamics. Multiple attempts have been made to approximate these dynamics to allow for an analytical control method; here four of these methods have been compared for controlling agile hopping, where there are large changes in forward velocity across a single stance. In addition, a new, empirical, approach has been demonstrated. In this, a simple control law is formulated, based on some simple approximations, which allows the parameters to be selected empirically through simulation. This has led to a controller able to offer similar performance to the best analytical approximation but with a much simpler form. This empirical controller may present new opportunities for controlling more complex dynamics and the development of a self-tuning method in future work.

UR - http://www.scopus.com/inward/record.url?scp=85026759487&partnerID=8YFLogxK

UR - http://dx.doi.org/10.1007/978-3-319-64107-2_7

U2 - 10.1007/978-3-319-64107-2_7

DO - 10.1007/978-3-319-64107-2_7

M3 - Conference contribution

SN - 9783319641065

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

SP - 79

EP - 85

BT - Towards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings

PB - Springer Verlag

CY - Germany

ER -