Papers by Lars Mikelsons

Title:	Modia and Julia for Grey Box Modeling
Authors:	Frederic Bruder and Lars Mikelsons
Abstract:	During the process of modelling an existing dynamic physical system, it may be hard to capture some of the phenomena exactly on the basis of only textbook-equations. With measurement data from the real system, approximators like artificial neural networks can help improve the models. However, simulation and machine learning are usually done in different software applications. A unified environment for modeling, simulation and optimization would be highly valuable. We here present a framework within the Julia programming language that encompasses tools for acausal modeling, automatic differentiation rsp. sensitivity analysis involving solvers for differential equations. We use it to build and evaluate an easily interpretable model based on both physics and data.
Keywords:	Grey Box Modeling, Hybrid Modeling, Scientific Machine Learning, Modia, Julia
Paper:	full paper
Bibtex:	@InProceedings{modelica.org:Bruder:2021, title = "{Modia and Julia for Grey Box Modeling}", author = {Frederic Bruder and Lars Mikelsons}, pages = {87--95}, doi = {10.3384/ecp2118187}, booktitle = {Proceedings of the 14th International Modelica Conference}, location = {Link\"oping, Sweden}, editor = {Martin Sj\"olund and Lena Buffoni and Adrian Pop and Lennart Ochel}, isbn = {978-91-7929-027-6}, issn = {1650-3740}, month = sep, series = {Link\"oping Electronic Conference Proceedings}, number = {181}, publisher = {Modelica Association and Link\"oping University Electronic Press}, year = {2021} }

Title:	NeuralFMU: Towards Structural Integration of FMUs into Neural Networks
Authors:	Tobias Thummerer, Josef Kircher and Lars Mikelsons
Abstract:	This paper covers two major subjects: First, the presentation of a new open-source library called FMI.jl for integrating FMI into the Julia programming environment by providing the possibility to load, parameterize and simulate FMUs. Further, an extension to this library called FMIFlux.jl is introduced, that allows the integration of FMUs into a neural network topology to obtain a NeuralFMU. This structural combination of an industry typical black-box model and a data-driven machine learning model combines the different advantages of both modeling approaches in one single development environment. This allows for the usage of advanced data driven modeling techniques for physical effects that are difficult to model based on first principles.
Keywords:	NeuralFMU, NeuralODE, FMI, FMU, Julia
Paper:	full paper
Bibtex:	@InProceedings{modelica.org:Thummerer:2021, title = "{NeuralFMU: Towards Structural Integration of FMUs into Neural Networks}", author = {Tobias Thummerer and Josef Kircher and Lars Mikelsons}, pages = {297--306}, doi = {10.3384/ecp21181297}, booktitle = {Proceedings of the 14th International Modelica Conference}, location = {Link\"oping, Sweden}, editor = {Martin Sj\"olund and Lena Buffoni and Adrian Pop and Lennart Ochel}, isbn = {978-91-7929-027-6}, issn = {1650-3740}, month = sep, series = {Link\"oping Electronic Conference Proceedings}, number = {181}, publisher = {Modelica Association and Link\"oping University Electronic Press}, year = {2021} }