Hybrid Neural Network Augmented Physics-based Models for Nonlinear Filtering

dc.contributor.authorImbiriba, Tales
dc.contributor.authorDemirkaya, Ahmet
dc.contributor.authorDuník, Jindřich
dc.contributor.authorStraka, Ondřej
dc.contributor.authorErdoğmuş, Deniz
dc.contributor.authorClosas, Pau
dc.date.accessioned2023-02-13T11:00:20Z
dc.date.available2023-02-13T11:00:20Z
dc.date.issued2022
dc.description.abstract-translatedIn this paper we present a hybrid neural network augmented physics-based modeling (APBM) framework for Bayesian nonlinear latent space estimation. The proposed APBM strategy allows for model adaptation when new operation conditions come into play or the physics-based model is insufficient (or incomplete) to properly describe the latent phenomenon. One advantage of the APBMs and our estimation procedure is the capability of maintaining the physical interpretability of estimated states. Furthermore, we propose a constraint filtering approach to control the neural network contributions to the overall model. We also exploit assumed density filtering techniques and cubature integration rules to present a flexible estimation strategy that can easily deal with nonlinear models and high-dimensional latent spaces. Finally, we demonstrate the efficacy of our methodology by leveraging a target tracking scenario with nonlinear and incomplete measurement and acceleration models, respectively.en
dc.format6 s.cs
dc.format.mimetypeapplication/pdf
dc.identifier.citationIMBIRIBA, T. DEMIRKAYA, A. DUNÍK, J. STRAKA, O. ERDOĞMUŞ, D. CLOSAS, P. Hybrid Neural Network Augmented Physics-based Models for Nonlinear Filtering. In Proceedings of the 25th International Conference on Information Fusion, FUSION 2022. Linköping, Sweden: IEEE, 2022. s. 1-6. ISBN: 978-1-73774-972-1 , ISSN: neuvedenocs
dc.identifier.document-number855689000065
dc.identifier.doi10.23919/FUSION49751.2022.9841291
dc.identifier.isbn978-1-73774-972-1
dc.identifier.issnneuvedeno
dc.identifier.obd43937078
dc.identifier.uri2-s2.0-85136554327
dc.identifier.urihttp://hdl.handle.net/11025/51459
dc.language.isoenen
dc.project.IDSGS-2022-022/Rozvoj a využití kybernetických systémů identifikace, diagnostiky a řízení 5cs
dc.publisherIEEEen
dc.relation.ispartofseriesProceedings of the 25th International Conference on Information Fusion, FUSION 2022en
dc.rightsPlný text je přístupný v rámci univerzity přihlášeným uživatelům.cs
dc.rights© IEEEen
dc.rights.accessrestrictedAccessen
dc.subject.translatedNonlinear filteringen
dc.subject.translatedTarget trackingen
dc.subject.translatedHybrid Neural Networken
dc.subject.translatedPhysics-based Neural Modelsen
dc.subject.translatedGaussian filteringen
dc.titleHybrid Neural Network Augmented Physics-based Models for Nonlinear Filteringen
dc.typekonferenční příspěvekcs
dc.typeConferenceObjecten
dc.type.statusPeer-revieweden
dc.type.versionpublishedVersionen

Files

Original bundle
Showing 1 - 1 out of 1 results
No Thumbnail Available
Name:
article_FUSION2022_ImDeDuStErCl.pdf
Size:
6.24 MB
Format:
Adobe Portable Document Format