Saved in:
Bibliographic Details
Main Authors: Brandt, Felix, Heuermann, Andreas, Hannebohm, Philip, Bachmann, Bernhard
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.09317
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911202545762304
author Brandt, Felix
Heuermann, Andreas
Hannebohm, Philip
Bachmann, Bernhard
author_facet Brandt, Felix
Heuermann, Andreas
Hannebohm, Philip
Bachmann, Bernhard
contents This paper presents a residual-informed machine learning approach for replacing algebraic loops in equation-based Modelica models with neural network surrogates. A feedforward neural network is trained using the residual (error) of the algebraic loop directly in its loss function, eliminating the need for a supervised dataset. This training strategy also resolves the issue of ambiguous solutions, allowing the surrogate to converge to a consistent solution rather than averaging multiple valid ones. Applied to the large-scale IEEE 14-Bus system, our method achieves a 60% reduction in simulation time compared to conventional simulations, while maintaining the same level of accuracy through error control mechanisms.
format Preprint
id arxiv_https___arxiv_org_abs_2510_09317
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Residual-Informed Learning of Solutions to Algebraic Loops
Brandt, Felix
Heuermann, Andreas
Hannebohm, Philip
Bachmann, Bernhard
Machine Learning
Numerical Analysis
65H10, 37M05
G.1.5; I.2.6
This paper presents a residual-informed machine learning approach for replacing algebraic loops in equation-based Modelica models with neural network surrogates. A feedforward neural network is trained using the residual (error) of the algebraic loop directly in its loss function, eliminating the need for a supervised dataset. This training strategy also resolves the issue of ambiguous solutions, allowing the surrogate to converge to a consistent solution rather than averaging multiple valid ones. Applied to the large-scale IEEE 14-Bus system, our method achieves a 60% reduction in simulation time compared to conventional simulations, while maintaining the same level of accuracy through error control mechanisms.
title Residual-Informed Learning of Solutions to Algebraic Loops
topic Machine Learning
Numerical Analysis
65H10, 37M05
G.1.5; I.2.6
url https://arxiv.org/abs/2510.09317