Improving transient learning behavior in model-free inversion-based iterative control with application to a desktop printer

Robin De Rozario; Tom Oomen

doi:10.1109/AMC.2019.8371136

Improving transient learning behavior in model-free inversion-based iterative control with application to a desktop printer

Robin De Rozario
, Tom Oomen

Control Systems Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

12 Citations (Scopus)

1 Downloads (Pure)

Abstract

Model-Free Inversion-based Iterative Control (MFIIC) enables tracking performance improvement of systems that perform repeating tasks without using a model of the system. The aim of this paper is (i) to show that MFIIC can result in a severe loss of performance if the Signal-to-Disturbance-Ratio (SDR) approaches 1, and (ii) to propose a solution to this problem. The Smoothed MFIIC (SMFIIC) method is developed, which does not suffer from the undesirable learning transient behavior. This is achieved by adaptively regulating the learning speed to ensure smooth convergence. The existence of bad learning transients in MFIIC and the efficacy of SMFIIC are illustrated on an experimental desktop printer.

Original language	English
Title of host publication	Proceedings - 2018 IEEE 15th International Workshop on Advanced Motion Control, AMC 2018
Publisher	Institute of Electrical and Electronics Engineers
Pages	455-460
Number of pages	6
ISBN (Electronic)	9781538619469
DOIs	https://doi.org/10.1109/AMC.2019.8371136
Publication status	Published - 1 Jun 2018
Event	15th IEEE International Workshop on Advanced Motion Control, AMC 2018 - Shibaura Institute of Technology, Tokyo, Japan Duration: 9 Mar 2018 → 11 Mar 2018 Conference number: 15 http://ewh.ieee.org/conf/amc/2018/

Conference

Conference	15th IEEE International Workshop on Advanced Motion Control, AMC 2018
Abbreviated title	AMC 2018
Country/Territory	Japan
City	Tokyo
Period	9/03/18 → 11/03/18
Other	AMC2018 is the 15th in a series of biennial workshops that brings together researchers active in the field of advanced motion control to discuss current developments and future perspectives on motion control technology and applications. The workshop will be held at Shibaura Institute of Technology, Tokyo, Japan, during March 9-11, 2018.
Internet address	http://ewh.ieee.org/conf/amc/2018/

Keywords

Frequency domain-analysis
Iterative learning control
Motion Control

Access to Document

10.1109/AMC.2019.8371136

Cite this

@inproceedings{40bf024276df48039d7e16d432aabcc4,

title = "Improving transient learning behavior in model-free inversion-based iterative control with application to a desktop printer",

abstract = "Model-Free Inversion-based Iterative Control (MFIIC) enables tracking performance improvement of systems that perform repeating tasks without using a model of the system. The aim of this paper is (i) to show that MFIIC can result in a severe loss of performance if the Signal-to-Disturbance-Ratio (SDR) approaches 1, and (ii) to propose a solution to this problem. The Smoothed MFIIC (SMFIIC) method is developed, which does not suffer from the undesirable learning transient behavior. This is achieved by adaptively regulating the learning speed to ensure smooth convergence. The existence of bad learning transients in MFIIC and the efficacy of SMFIIC are illustrated on an experimental desktop printer.",

keywords = "Frequency domain-analysis, Iterative learning control, Motion Control",

author = "\{De Rozario\}, Robin and Tom Oomen",

year = "2018",

month = jun,

day = "1",

doi = "10.1109/AMC.2019.8371136",

language = "English",

pages = "455--460",

booktitle = "Proceedings - 2018 IEEE 15th International Workshop on Advanced Motion Control, AMC 2018",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

note = "15th IEEE International Workshop on Advanced Motion Control, AMC 2018, AMC 2018 ; Conference date: 09-03-2018 Through 11-03-2018",

url = "http://ewh.ieee.org/conf/amc/2018/",

}

De Rozario, R & Oomen, T 2018, Improving transient learning behavior in model-free inversion-based iterative control with application to a desktop printer. in Proceedings - 2018 IEEE 15th International Workshop on Advanced Motion Control, AMC 2018. Institute of Electrical and Electronics Engineers, pp. 455-460, 15th IEEE International Workshop on Advanced Motion Control, AMC 2018, Tokyo, Japan, 9/03/18. https://doi.org/10.1109/AMC.2019.8371136

Improving transient learning behavior in model-free inversion-based iterative control with application to a desktop printer. / De Rozario, Robin; Oomen, Tom.
Proceedings - 2018 IEEE 15th International Workshop on Advanced Motion Control, AMC 2018. Institute of Electrical and Electronics Engineers, 2018. p. 455-460.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Improving transient learning behavior in model-free inversion-based iterative control with application to a desktop printer

AU - De Rozario, Robin

AU - Oomen, Tom

N1 - Conference code: 15

PY - 2018/6/1

Y1 - 2018/6/1

N2 - Model-Free Inversion-based Iterative Control (MFIIC) enables tracking performance improvement of systems that perform repeating tasks without using a model of the system. The aim of this paper is (i) to show that MFIIC can result in a severe loss of performance if the Signal-to-Disturbance-Ratio (SDR) approaches 1, and (ii) to propose a solution to this problem. The Smoothed MFIIC (SMFIIC) method is developed, which does not suffer from the undesirable learning transient behavior. This is achieved by adaptively regulating the learning speed to ensure smooth convergence. The existence of bad learning transients in MFIIC and the efficacy of SMFIIC are illustrated on an experimental desktop printer.

AB - Model-Free Inversion-based Iterative Control (MFIIC) enables tracking performance improvement of systems that perform repeating tasks without using a model of the system. The aim of this paper is (i) to show that MFIIC can result in a severe loss of performance if the Signal-to-Disturbance-Ratio (SDR) approaches 1, and (ii) to propose a solution to this problem. The Smoothed MFIIC (SMFIIC) method is developed, which does not suffer from the undesirable learning transient behavior. This is achieved by adaptively regulating the learning speed to ensure smooth convergence. The existence of bad learning transients in MFIIC and the efficacy of SMFIIC are illustrated on an experimental desktop printer.

KW - Frequency domain-analysis

KW - Iterative learning control

KW - Motion Control

UR - https://www.scopus.com/pages/publications/85048780029

U2 - 10.1109/AMC.2019.8371136

DO - 10.1109/AMC.2019.8371136

M3 - Conference contribution

AN - SCOPUS:85048780029

SP - 455

EP - 460

BT - Proceedings - 2018 IEEE 15th International Workshop on Advanced Motion Control, AMC 2018

PB - Institute of Electrical and Electronics Engineers

T2 - 15th IEEE International Workshop on Advanced Motion Control, AMC 2018

Y2 - 9 March 2018 through 11 March 2018

ER -

Improving transient learning behavior in model-free inversion-based iterative control with application to a desktop printer

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this