Cargando…

Intelliquench: An Adaptive Machine Learning System for Detection of Superconducting Magnet Quenches

In superconducting magnets, the irreversible transition of a portion of the conductor to resistive state is called a “quench.” Having large stored energy, magnets can be damaged by quenches due to localized heating, high voltage, or large force transients. Unfortunately, current quench protection sy...

Descripción completa

Detalles Bibliográficos
Autores principales: Hoang, Duc, Boffo, Cristian, Tran, Nhan, Krave, Steven, Kazi, Sujay, Stoynev, Stoyan, Marinozzi, Vittorio
Lenguaje:eng
Publicado: 2021
Materias:
Acceso en línea:https://dx.doi.org/10.1109/TASC.2021.3058229
http://cds.cern.ch/record/2770758
_version_ 1780971400006205440
author Hoang, Duc
Boffo, Cristian
Tran, Nhan
Krave, Steven
Kazi, Sujay
Stoynev, Stoyan
Marinozzi, Vittorio
author_facet Hoang, Duc
Boffo, Cristian
Tran, Nhan
Krave, Steven
Kazi, Sujay
Stoynev, Stoyan
Marinozzi, Vittorio
author_sort Hoang, Duc
collection CERN
description In superconducting magnets, the irreversible transition of a portion of the conductor to resistive state is called a “quench.” Having large stored energy, magnets can be damaged by quenches due to localized heating, high voltage, or large force transients. Unfortunately, current quench protection systems can only detect a quench after it happens, and mitigating risks in Low Temperature Superconducting (LTS) accelerator magnets often requires fast response (down to ms). Additionally, protection of High Temperature Superconducting (HTS) magnets is still suffering from prohibitively slow quench detection. In this study, we lay the groundwork for a quench prediction system using an auto-encoder fully-connected deep neural network. After dynamically trained with data features extracted from acoustic sensors around the magnet, the system detects anomalous events seconds before the quench in most of our data. While the exact nature of the events is under investigation, we show that the system can “forecast” a quench before it happens under magnet training conditions through a randomized experiment. This opens up the way of integrated data processing, potentially leading to faster and better diagnostics and detection of magnet quenches
id oai-inspirehep.net-1845760
institution Organización Europea para la Investigación Nuclear
language eng
publishDate 2021
record_format invenio
spelling oai-inspirehep.net-18457602021-05-31T15:21:44Zdoi:10.1109/TASC.2021.3058229http://cds.cern.ch/record/2770758engHoang, DucBoffo, CristianTran, NhanKrave, StevenKazi, SujayStoynev, StoyanMarinozzi, VittorioIntelliquench: An Adaptive Machine Learning System for Detection of Superconducting Magnet QuenchesAccelerators and Storage RingsIn superconducting magnets, the irreversible transition of a portion of the conductor to resistive state is called a “quench.” Having large stored energy, magnets can be damaged by quenches due to localized heating, high voltage, or large force transients. Unfortunately, current quench protection systems can only detect a quench after it happens, and mitigating risks in Low Temperature Superconducting (LTS) accelerator magnets often requires fast response (down to ms). Additionally, protection of High Temperature Superconducting (HTS) magnets is still suffering from prohibitively slow quench detection. In this study, we lay the groundwork for a quench prediction system using an auto-encoder fully-connected deep neural network. After dynamically trained with data features extracted from acoustic sensors around the magnet, the system detects anomalous events seconds before the quench in most of our data. While the exact nature of the events is under investigation, we show that the system can “forecast” a quench before it happens under magnet training conditions through a randomized experiment. This opens up the way of integrated data processing, potentially leading to faster and better diagnostics and detection of magnet quenchesIn superconducting magnets, the irreversible transition of a portion of the conductor to resistive state is called a “quench.” Having large stored energy, magnets can be damaged by quenches due to localized heating, high voltage, or large force transients. Unfortunately, current quench protection systems can only detect a quench after it happens, and mitigating risks in Low Temperature Superconducting (LTS) accelerator magnets often requires fast response (down to ms). Additionally, protection of High Temperature Superconducting (HTS) magnets is still suffering from prohibitively slow quench detection. In this study, we lay the groundwork for a quench prediction system using an auto-encoder fully-connected deep neural network. After dynamically trained with data features extracted from acoustic sensors around the magnet, the system detects anomalous events seconds before the quench in most of our data. While the exact nature of the events is under investigation, we show that the system can “forecast” a quench before it happens under magnet training conditions through a randomized experiment. This opens up the way of integrated data processing, potentially leading to faster and better diagnostics and detection of magnet quenches.FERMILAB-PUB-21-035-CMS-SCD-STUDENT-TDoai:inspirehep.net:18457602021
spellingShingle Accelerators and Storage Rings
Hoang, Duc
Boffo, Cristian
Tran, Nhan
Krave, Steven
Kazi, Sujay
Stoynev, Stoyan
Marinozzi, Vittorio
Intelliquench: An Adaptive Machine Learning System for Detection of Superconducting Magnet Quenches
title Intelliquench: An Adaptive Machine Learning System for Detection of Superconducting Magnet Quenches
title_full Intelliquench: An Adaptive Machine Learning System for Detection of Superconducting Magnet Quenches
title_fullStr Intelliquench: An Adaptive Machine Learning System for Detection of Superconducting Magnet Quenches
title_full_unstemmed Intelliquench: An Adaptive Machine Learning System for Detection of Superconducting Magnet Quenches
title_short Intelliquench: An Adaptive Machine Learning System for Detection of Superconducting Magnet Quenches
title_sort intelliquench: an adaptive machine learning system for detection of superconducting magnet quenches
topic Accelerators and Storage Rings
url https://dx.doi.org/10.1109/TASC.2021.3058229
http://cds.cern.ch/record/2770758
work_keys_str_mv AT hoangduc intelliquenchanadaptivemachinelearningsystemfordetectionofsuperconductingmagnetquenches
AT boffocristian intelliquenchanadaptivemachinelearningsystemfordetectionofsuperconductingmagnetquenches
AT trannhan intelliquenchanadaptivemachinelearningsystemfordetectionofsuperconductingmagnetquenches
AT kravesteven intelliquenchanadaptivemachinelearningsystemfordetectionofsuperconductingmagnetquenches
AT kazisujay intelliquenchanadaptivemachinelearningsystemfordetectionofsuperconductingmagnetquenches
AT stoynevstoyan intelliquenchanadaptivemachinelearningsystemfordetectionofsuperconductingmagnetquenches
AT marinozzivittorio intelliquenchanadaptivemachinelearningsystemfordetectionofsuperconductingmagnetquenches