Optimal Control of Multi-phase Movements with Learned Dynamics

Andreea Radulescu; Jun Nakanishi; Sethu Vijayakumar

doi:10.1007/978-3-319-23437-3_5

Optimal Control of Multi-phase Movements with Learned Dynamics

Andreea Radulescu, Jun Nakanishi, Sethu Vijayakumar

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract / Description of output

In this paper, we extend our work on movement optimisation for variable stiffness actuation (VSA) with multiple phases and switching dynamics to incorporate scenarios with incomplete, complex or hard to model robot dynamics. By incorporating a locally weighted nonparametric learning method to model the discrepancies in the system dynamics, we formulate an online adaptation scheme capable of systematically improving the multi-phase plans (stiffness modulation and torques) and switching instances while improving the dynamics model on the fly. This is demonstrated on a realistic model of a VSA brachiating system with excellent adaptation results.

Original language	English
Title of host publication	Man–Machine Interactions 4
Subtitle of host publication	4th International Conference on Man–Machine Interactions, ICMMI 2015 Kocierz Pass, Poland, October 6–9, 2015
Editors	Aleksandra Gruca, Agnieszka Brachman, Stanislaw Kozielski, Tadeusz Czachórski
Place of Publication	Cham
Publisher	Springer
Pages	61-76
Number of pages	16
ISBN (Electronic)	978-3-319-23437-3
ISBN (Print)	978-3-319-23436-6
DOIs	https://doi.org/10.1007/978-3-319-23437-3_5
Publication status	Published - 9 Sept 2016
Event	4th International Conference on Man–Machine Interactions - Kocierz Pass, Poland Duration: 6 Oct 2015 → 9 Oct 2015

Publication series

Name	Advances in Intelligent Systems and Computing
Publisher	Springer International Publishing
Volume	391
ISSN (Print)	2194-5357

Conference

Conference	4th International Conference on Man–Machine Interactions
Abbreviated title	ICMMI 2015
Country/Territory	Poland
City	Kocierz Pass
Period	6/10/15 → 9/10/15

Access to Document

10.1007/978-3-319-23437-3_5

ICMMI_2015-1Accepted author manuscript, 1.96 MB

http://link.springer.com/chapter/10.1007/978-3-319-23437-3_5

Cite this

Radulescu, A., Nakanishi, J., & Vijayakumar, S. (2016). Optimal Control of Multi-phase Movements with Learned Dynamics. In A. Gruca, A. Brachman, S. Kozielski, & T. Czachórski (Eds.), Man–Machine Interactions 4: 4th International Conference on Man–Machine Interactions, ICMMI 2015 Kocierz Pass, Poland, October 6–9, 2015 (pp. 61-76). (Advances in Intelligent Systems and Computing; Vol. 391). Springer. https://doi.org/10.1007/978-3-319-23437-3_5

Radulescu, Andreea ; Nakanishi, Jun ; Vijayakumar, Sethu. / Optimal Control of Multi-phase Movements with Learned Dynamics. Man–Machine Interactions 4: 4th International Conference on Man–Machine Interactions, ICMMI 2015 Kocierz Pass, Poland, October 6–9, 2015. editor / Aleksandra Gruca ; Agnieszka Brachman ; Stanislaw Kozielski ; Tadeusz Czachórski. Cham : Springer, 2016. pp. 61-76 (Advances in Intelligent Systems and Computing).

@inproceedings{0c2b940b275349009a90e540037724b2,

title = "Optimal Control of Multi-phase Movements with Learned Dynamics",

abstract = "In this paper, we extend our work on movement optimisation for variable stiffness actuation (VSA) with multiple phases and switching dynamics to incorporate scenarios with incomplete, complex or hard to model robot dynamics. By incorporating a locally weighted nonparametric learning method to model the discrepancies in the system dynamics, we formulate an online adaptation scheme capable of systematically improving the multi-phase plans (stiffness modulation and torques) and switching instances while improving the dynamics model on the fly. This is demonstrated on a realistic model of a VSA brachiating system with excellent adaptation results.",

author = "Andreea Radulescu and Jun Nakanishi and Sethu Vijayakumar",

year = "2016",

month = sep,

day = "9",

doi = "10.1007/978-3-319-23437-3_5",

language = "English",

isbn = "978-3-319-23436-6",

series = "Advances in Intelligent Systems and Computing",

publisher = "Springer",

pages = "61--76",

editor = "Aleksandra Gruca and Agnieszka Brachman and Stanislaw Kozielski and Tadeusz Czach{\'o}rski",

booktitle = "Man–Machine Interactions 4",

address = "United Kingdom",

note = "4th International Conference on Man–Machine Interactions, ICMMI 2015 ; Conference date: 06-10-2015 Through 09-10-2015",

}

Radulescu, A, Nakanishi, J & Vijayakumar, S 2016, Optimal Control of Multi-phase Movements with Learned Dynamics. in A Gruca, A Brachman, S Kozielski & T Czachórski (eds), Man–Machine Interactions 4: 4th International Conference on Man–Machine Interactions, ICMMI 2015 Kocierz Pass, Poland, October 6–9, 2015. Advances in Intelligent Systems and Computing, vol. 391, Springer, Cham, pp. 61-76, 4th International Conference on Man–Machine Interactions, Kocierz Pass, Poland, 6/10/15. https://doi.org/10.1007/978-3-319-23437-3_5

Optimal Control of Multi-phase Movements with Learned Dynamics. / Radulescu, Andreea; Nakanishi, Jun; Vijayakumar, Sethu.
Man–Machine Interactions 4: 4th International Conference on Man–Machine Interactions, ICMMI 2015 Kocierz Pass, Poland, October 6–9, 2015. ed. / Aleksandra Gruca; Agnieszka Brachman; Stanislaw Kozielski; Tadeusz Czachórski. Cham: Springer, 2016. p. 61-76 (Advances in Intelligent Systems and Computing; Vol. 391).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Optimal Control of Multi-phase Movements with Learned Dynamics

AU - Radulescu, Andreea

AU - Nakanishi, Jun

AU - Vijayakumar, Sethu

PY - 2016/9/9

Y1 - 2016/9/9

N2 - In this paper, we extend our work on movement optimisation for variable stiffness actuation (VSA) with multiple phases and switching dynamics to incorporate scenarios with incomplete, complex or hard to model robot dynamics. By incorporating a locally weighted nonparametric learning method to model the discrepancies in the system dynamics, we formulate an online adaptation scheme capable of systematically improving the multi-phase plans (stiffness modulation and torques) and switching instances while improving the dynamics model on the fly. This is demonstrated on a realistic model of a VSA brachiating system with excellent adaptation results.

AB - In this paper, we extend our work on movement optimisation for variable stiffness actuation (VSA) with multiple phases and switching dynamics to incorporate scenarios with incomplete, complex or hard to model robot dynamics. By incorporating a locally weighted nonparametric learning method to model the discrepancies in the system dynamics, we formulate an online adaptation scheme capable of systematically improving the multi-phase plans (stiffness modulation and torques) and switching instances while improving the dynamics model on the fly. This is demonstrated on a realistic model of a VSA brachiating system with excellent adaptation results.

U2 - 10.1007/978-3-319-23437-3_5

DO - 10.1007/978-3-319-23437-3_5

M3 - Conference contribution

SN - 978-3-319-23436-6

T3 - Advances in Intelligent Systems and Computing

SP - 61

EP - 76

BT - Man–Machine Interactions 4

A2 - Gruca, Aleksandra

A2 - Brachman, Agnieszka

A2 - Kozielski, Stanislaw

A2 - Czachórski, Tadeusz

PB - Springer

CY - Cham

T2 - 4th International Conference on Man–Machine Interactions

Y2 - 6 October 2015 through 9 October 2015

ER -

Radulescu A, Nakanishi J, Vijayakumar S. Optimal Control of Multi-phase Movements with Learned Dynamics. In Gruca A, Brachman A, Kozielski S, Czachórski T, editors, Man–Machine Interactions 4: 4th International Conference on Man–Machine Interactions, ICMMI 2015 Kocierz Pass, Poland, October 6–9, 2015. Cham: Springer. 2016. p. 61-76. (Advances in Intelligent Systems and Computing). doi: 10.1007/978-3-319-23437-3_5

Optimal Control of Multi-phase Movements with Learned Dynamics

Abstract / Description of output

Publication series

Conference

Access to Document

Fingerprint

Cite this