Gaussian Process based Feedforward Control for Nonlinear Systems with Flexible Tasks: With Application to a Printer with Friction

Max Van Meer; Maurice Poot; Jim Portegies; Tom Oomen

doi:10.1016/j.ifacol.2022.11.191

Gaussian Process based Feedforward Control for Nonlinear Systems with Flexible Tasks: With Application to a Printer with Friction

Max Van Meer, Maurice Poot, Jim Portegies, Tom Oomen

Team Jan-Willem van Wingerden

Research output: Contribution to journal › Conference article › Scientific › peer-review

2 Citations (Scopus)

72 Downloads (Pure)

Abstract

Feedforward control is essential to achieving good tracking performance in positioning systems. The aim of this paper is to develop an identification strategy for inverse models of systems with nonlinear dynamics of unknown structure using input-output data, which can be used to generate feedforward signals for a-priori unknown tasks. To this end, inverse systems are regarded as noncausal nonlinear finite impulse response (NFIR) systems, and modeled as a Gaussian Process with a stationary kernel function that imposes properties such as smoothness. The approach is validated experimentally on a consumer printer with friction and shown to lead to improved tracking performance with respect to linear feedforward.

Original language	English
Pages (from-to)	241-246
Journal	IFAC-PapersOnline
Volume	55
Issue number	37
DOIs	https://doi.org/10.1016/j.ifacol.2022.11.191
Publication status	Published - 2022
Event	2nd Modeling, Estimation and Control Conference, MECC 2022 - Jersey City, United States Duration: 2 Oct 2022 → 5 Oct 2022

Keywords

Feedforward Control
Gaussian Process regression
Grey box modelling
Identification for control
Nonlinear system identification

Access to Document

10.1016/j.ifacol.2022.11.191

1-s2.0-S2405896322028348-mainFinal published version, 1.74 MBLicence: CC BY-NC-ND

Cite this

@article{0843a9a2b91248498b72ce4f541ca259,

title = "Gaussian Process based Feedforward Control for Nonlinear Systems with Flexible Tasks: With Application to a Printer with Friction",

abstract = "Feedforward control is essential to achieving good tracking performance in positioning systems. The aim of this paper is to develop an identification strategy for inverse models of systems with nonlinear dynamics of unknown structure using input-output data, which can be used to generate feedforward signals for a-priori unknown tasks. To this end, inverse systems are regarded as noncausal nonlinear finite impulse response (NFIR) systems, and modeled as a Gaussian Process with a stationary kernel function that imposes properties such as smoothness. The approach is validated experimentally on a consumer printer with friction and shown to lead to improved tracking performance with respect to linear feedforward.",

keywords = "Feedforward Control, Gaussian Process regression, Grey box modelling, Identification for control, Nonlinear system identification",

author = "{Van Meer}, Max and Maurice Poot and Jim Portegies and Tom Oomen",

year = "2022",

doi = "10.1016/j.ifacol.2022.11.191",

language = "English",

volume = "55",

pages = "241--246",

journal = "IFAC-PapersOnline",

issn = "1474-6670",

publisher = "Elsevier",

number = "37",

note = "2nd Modeling, Estimation and Control Conference, MECC 2022 ; Conference date: 02-10-2022 Through 05-10-2022",

}

TY - JOUR

T1 - Gaussian Process based Feedforward Control for Nonlinear Systems with Flexible Tasks

T2 - 2nd Modeling, Estimation and Control Conference, MECC 2022

AU - Van Meer, Max

AU - Poot, Maurice

AU - Portegies, Jim

AU - Oomen, Tom

PY - 2022

Y1 - 2022

N2 - Feedforward control is essential to achieving good tracking performance in positioning systems. The aim of this paper is to develop an identification strategy for inverse models of systems with nonlinear dynamics of unknown structure using input-output data, which can be used to generate feedforward signals for a-priori unknown tasks. To this end, inverse systems are regarded as noncausal nonlinear finite impulse response (NFIR) systems, and modeled as a Gaussian Process with a stationary kernel function that imposes properties such as smoothness. The approach is validated experimentally on a consumer printer with friction and shown to lead to improved tracking performance with respect to linear feedforward.

AB - Feedforward control is essential to achieving good tracking performance in positioning systems. The aim of this paper is to develop an identification strategy for inverse models of systems with nonlinear dynamics of unknown structure using input-output data, which can be used to generate feedforward signals for a-priori unknown tasks. To this end, inverse systems are regarded as noncausal nonlinear finite impulse response (NFIR) systems, and modeled as a Gaussian Process with a stationary kernel function that imposes properties such as smoothness. The approach is validated experimentally on a consumer printer with friction and shown to lead to improved tracking performance with respect to linear feedforward.

KW - Feedforward Control

KW - Gaussian Process regression

KW - Grey box modelling

KW - Identification for control

KW - Nonlinear system identification

UR - http://www.scopus.com/inward/record.url?scp=85146147160&partnerID=8YFLogxK

U2 - 10.1016/j.ifacol.2022.11.191

DO - 10.1016/j.ifacol.2022.11.191

M3 - Conference article

AN - SCOPUS:85146147160

SN - 1474-6670

VL - 55

SP - 241

EP - 246

JO - IFAC-PapersOnline

JF - IFAC-PapersOnline

IS - 37

Y2 - 2 October 2022 through 5 October 2022

ER -

Gaussian Process based Feedforward Control for Nonlinear Systems with Flexible Tasks: With Application to a Printer with Friction

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this