Long short-term memory for improved transients in neural network adaptive control

Simple item page
buir.contributor.authorİnanç, Emirhan
buir.contributor.authorGürses, Yiğit
buir.contributor.authorHabboush, Abdullah
buir.contributor.authorYıldız, Yıldıray
buir.contributor.orcidİnanç, Emirhan|0009-0002-7541-960X
buir.contributor.orcidGürses, Yiğit|0009-0008-3367-7495
buir.contributor.orcidHabboush, Abdullah|0000-0001-8598-6419
buir.contributor.orcidYıldız, Yıldıray|0000-0001-6270-5354
dc.citation.epage3619en_US
dc.citation.spage3614
dc.contributor.authorİnanç, Emirhan
dc.contributor.authorGürses, Yiğit
dc.contributor.authorHabboush, Abdullah
dc.contributor.authorYıldız, Yıldıray
dc.coverage.spatialSan Diego, CA, USA
dc.date.accessioned2024-03-08T09:09:35Z
dc.date.available2024-03-08T09:09:35Z
dc.date.issued2023-07-03
dc.departmentDepartment of Mechanical Engineering
dc.departmentDepartment of Computer Engineering
dc.descriptionConference Name: 2023 American Control Conference (ACC)
dc.descriptionDate of Conference: 31 May 2023 - 02 June 2023
dc.description.abstractIn this study, we propose a novel adaptive control architecture, which provides dramatically better performance compared to conventional methods. What makes this architecture unique is the synergistic employment of a traditional, Adaptive Neural Network (ANN) controller and a Long Short-Term Memory (LSTM) network. LSTM structures, unlike the standard feed-forward neural networks, take advantage of the dependencies in an input sequence, which helps predict the evolution of an uncertainty. Through a training method we introduced, the LSTM network learns to compensate for the deficiencies of the ANN controller. This substantially improves the transient response by allowing the controller to quickly react to unexpected events. Through careful simulation studies, we demonstrate that this architecture can improve the estimation accuracy on a diverse set of unseen uncertainties. We also provide an analysis of the contributions of the ANN controller and LSTM network, identifying their individual roles in compensating low and high frequency error dynamics. This analysis provides insight into why and how the LSTM augmentation improves the system’s transient response.
dc.identifier.doi10.23919/ACC55779.2023.10155805
dc.identifier.eisbn979-8-3503-2806-6
dc.identifier.eissn2378-5861
dc.identifier.isbn978-1-6654-6952-4
dc.identifier.isbn979-8-3503-2807-3
dc.identifier.issn0743-1619
dc.identifier.urihttps://hdl.handle.net/11693/114407
dc.language.isoen_US
dc.publisherIEEE
dc.relation.isversionofhttps://dx.doi.org/10.23919/ACC55779.2023.10155805
dc.source.title2023 American Control Conference (ACC)
dc.subjectTraining
dc.subjectTransient response
dc.subjectUncertainty
dc.subjectEstimation
dc.subjectFeedforward neural networks
dc.subjectHigh frequency
dc.subjectAdaptive control
dc.titleLong short-term memory for improved transients in neural network adaptive control
dc.typeConference Paper

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Long_short-term_memory_for_improved_transients_in_neural_network_adaptive_control.pdf
Size:
2.22 MB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
2.01 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

Department of Mechanical Engineering
Department of Computer Engineering