Employing automatic differentiation and neural networks for parameter identification of an energy based hysteresis model

Eniz Museljic; Klaus Roppert; Lukas Daniel Domenig; Alice Reinbacher-Köstinger; Manfred Kaltenbacher

doi:10.3233/JAE-230107

Employing automatic differentiation and neural networks for parameter identification of an energy based hysteresis model

Eniz Museljic, Klaus Roppert, Lukas Daniel Domenig, Alice Reinbacher-Köstinger, Manfred Kaltenbacher

Institute of Fundamentals and Theory in Electrical Engineering (4370)

Research output: Contribution to journal › Article › peer-review

Abstract

This paper is about the parameter identification of an energy based hysteresis model from measurements by employing automatic differentiation and neural networks. We first introduce the energy based hysteresis model and the parameters which are to be identified. Then we show how the model can benefit from automatic differentiation. After that we incorporate a parametrization of the energy based hysteresis model via distribution functions and identify the parameters of the distribution function. Then, the hysteresis model is sampled and the generated datasets are used to train neural networks to predict the hysteresis parameters. The described methods are tested and verified on synthetic as well as measurement data.

Original language	English
Pages (from-to)	415-427
Number of pages	13
Journal	International Journal of Applied Electromagnetics and Mechanics
Volume	73
Issue number	4
DOIs	https://doi.org/10.3233/JAE-230107
Publication status	Published - 14 Dec 2023

Keywords

Parameter identification
Hysteresis modeling
Neural network
Optimization
neural networks
machine learning
hysteresis
parameter identification

ASJC Scopus subject areas

Electronic, Optical and Magnetic Materials
Condensed Matter Physics
Mechanics of Materials
Mechanical Engineering
Electrical and Electronic Engineering

Access to Document

10.3233/JAE-230107

Cite this

@article{a9e90811926b4d2a97ef900d1eda40b0,

title = "Employing automatic differentiation and neural networks for parameter identification of an energy based hysteresis model",

abstract = "This paper is about the parameter identification of an energy based hysteresis model from measurements by employing automatic differentiation and neural networks. We first introduce the energy based hysteresis model and the parameters which are to be identified. Then we show how the model can benefit from automatic differentiation. After that we incorporate a parametrization of the energy based hysteresis model via distribution functions and identify the parameters of the distribution function. Then, the hysteresis model is sampled and the generated datasets are used to train neural networks to predict the hysteresis parameters. The described methods are tested and verified on synthetic as well as measurement data.",

keywords = "Parameter identification, Hysteresis modeling, Neural network, Optimization, neural networks, machine learning, hysteresis, parameter identification",

author = "Eniz Museljic and Klaus Roppert and Domenig, {Lukas Daniel} and Alice Reinbacher-K{\"o}stinger and Manfred Kaltenbacher",

year = "2023",

month = dec,

day = "14",

doi = "10.3233/JAE-230107",

language = "English",

volume = "73",

pages = "415--427",

journal = "International Journal of Applied Electromagnetics and Mechanics",

issn = "1383-5416",

publisher = "IOS Press",

number = "4",

}

TY - JOUR

T1 - Employing automatic differentiation and neural networks for parameter identification of an energy based hysteresis model

AU - Museljic, Eniz

AU - Roppert, Klaus

AU - Domenig, Lukas Daniel

AU - Reinbacher-Köstinger, Alice

AU - Kaltenbacher, Manfred

PY - 2023/12/14

Y1 - 2023/12/14

N2 - This paper is about the parameter identification of an energy based hysteresis model from measurements by employing automatic differentiation and neural networks. We first introduce the energy based hysteresis model and the parameters which are to be identified. Then we show how the model can benefit from automatic differentiation. After that we incorporate a parametrization of the energy based hysteresis model via distribution functions and identify the parameters of the distribution function. Then, the hysteresis model is sampled and the generated datasets are used to train neural networks to predict the hysteresis parameters. The described methods are tested and verified on synthetic as well as measurement data.

AB - This paper is about the parameter identification of an energy based hysteresis model from measurements by employing automatic differentiation and neural networks. We first introduce the energy based hysteresis model and the parameters which are to be identified. Then we show how the model can benefit from automatic differentiation. After that we incorporate a parametrization of the energy based hysteresis model via distribution functions and identify the parameters of the distribution function. Then, the hysteresis model is sampled and the generated datasets are used to train neural networks to predict the hysteresis parameters. The described methods are tested and verified on synthetic as well as measurement data.

KW - Parameter identification

KW - Hysteresis modeling

KW - Neural network

KW - Optimization

KW - neural networks

KW - machine learning

KW - hysteresis

KW - parameter identification

UR - http://www.scopus.com/inward/record.url?scp=85180267797&partnerID=8YFLogxK

U2 - 10.3233/JAE-230107

DO - 10.3233/JAE-230107

M3 - Article

SN - 1383-5416

VL - 73

SP - 415

EP - 427

JO - International Journal of Applied Electromagnetics and Mechanics

JF - International Journal of Applied Electromagnetics and Mechanics

IS - 4

ER -

Employing automatic differentiation and neural networks for parameter identification of an energy based hysteresis model

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this