A Factor Graph Approach to Parameter Identification for Affine LPV Systems

C. Hoffmann; A. Isler; P. Rostalski

doi:10.23919/ACC.2017.7963231

A Factor Graph Approach to Parameter Identification for Affine LPV Systems

C. Hoffmann, A. Isler, P. Rostalski

Institute of Electrical Engineering in Medicine

ABB Switzerland Ltd

Abstract

Factor graphs are a versatile graphical representation of factorizable functions. As a probabilistic graphical model they allow to visualize conditional independence, which can be exploited for efficiently solving inference problems by means of message passing along the nodes of the graph. In this paper, a novel factor graph formulation of the expectation maximization-based estimation technique for affine linear parameter-varying system identification is presented. Furthermore, a recursive reformulation of the algorithm suitable for tracking time-varying changes both accounted and unaccounted for by a pre-defined linear parameter-varying system description is immediate from its factor graph-based formulation.

Original language	English
Title of host publication	2017 American Control Conference (ACC)
Number of pages	6
Publisher	IEEE
Publication date	29.06.2017
Pages	1910-1915
Article number	7963231
ISBN (Print)	978-1-5090-5994-2, 978-1-5090-4583-9
ISBN (Electronic)	978-1-5090-5992-8
DOIs	https://doi.org/10.23919/ACC.2017.7963231
Publication status	Published - 29.06.2017
Event	2017 American Control Conference - Sheraton Seattle Hotel , Seattle, United States Duration: 24.05.2017 → 26.05.2017 Conference number: 128855

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.23919/ACC.2017.7963231

Cite this

@inproceedings{c187b3f3c7b24d7faf2df09721adc6e8,

title = "A Factor Graph Approach to Parameter Identification for Affine LPV Systems",

abstract = "Factor graphs are a versatile graphical representation of factorizable functions. As a probabilistic graphical model they allow to visualize conditional independence, which can be exploited for efficiently solving inference problems by means of message passing along the nodes of the graph. In this paper, a novel factor graph formulation of the expectation maximization-based estimation technique for affine linear parameter-varying system identification is presented. Furthermore, a recursive reformulation of the algorithm suitable for tracking time-varying changes both accounted and unaccounted for by a pre-defined linear parameter-varying system description is immediate from its factor graph-based formulation.",

author = "C. Hoffmann and A. Isler and P. Rostalski",

year = "2017",

month = jun,

day = "29",

doi = "10.23919/ACC.2017.7963231",

language = "English",

isbn = "978-1-5090-5994-2",

series = "Proceedings of the American Control Conference",

publisher = "IEEE",

pages = "1910--1915",

booktitle = "2017 American Control Conference (ACC)",

note = "2017 American Control Conference<br/>, ACC 2017 ; Conference date: 24-05-2017 Through 26-05-2017",

}

TY - GEN

T1 - A Factor Graph Approach to Parameter Identification for Affine LPV Systems

AU - Hoffmann, C.

AU - Isler, A.

AU - Rostalski, P.

N1 - Conference code: 128855

PY - 2017/6/29

Y1 - 2017/6/29

N2 - Factor graphs are a versatile graphical representation of factorizable functions. As a probabilistic graphical model they allow to visualize conditional independence, which can be exploited for efficiently solving inference problems by means of message passing along the nodes of the graph. In this paper, a novel factor graph formulation of the expectation maximization-based estimation technique for affine linear parameter-varying system identification is presented. Furthermore, a recursive reformulation of the algorithm suitable for tracking time-varying changes both accounted and unaccounted for by a pre-defined linear parameter-varying system description is immediate from its factor graph-based formulation.

AB - Factor graphs are a versatile graphical representation of factorizable functions. As a probabilistic graphical model they allow to visualize conditional independence, which can be exploited for efficiently solving inference problems by means of message passing along the nodes of the graph. In this paper, a novel factor graph formulation of the expectation maximization-based estimation technique for affine linear parameter-varying system identification is presented. Furthermore, a recursive reformulation of the algorithm suitable for tracking time-varying changes both accounted and unaccounted for by a pre-defined linear parameter-varying system description is immediate from its factor graph-based formulation.

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-85027032211&origin=inward&txGid=5b43807028f63968de7ae1ce93fbe6ea

U2 - 10.23919/ACC.2017.7963231

DO - 10.23919/ACC.2017.7963231

M3 - Conference contribution

SN - 978-1-5090-5994-2

SN - 978-1-5090-4583-9

T3 - Proceedings of the American Control Conference

SP - 1910

EP - 1915

BT - 2017 American Control Conference (ACC)

PB - IEEE

T2 - 2017 American Control Conference<br/>

Y2 - 24 May 2017 through 26 May 2017

ER -

A Factor Graph Approach to Parameter Identification for Affine LPV Systems

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this