A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators
The beam interruptions (interlocks) of particle accelerators, despite being necessary safety measures, lead to abrupt operational changes and a substantial loss of beam time. A novel time series classification approach is applied to decrease beam time loss in the High-Intensity Proton Accelerator co...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-03-01
|
Series: | Information |
Subjects: | |
Online Access: | https://www.mdpi.com/2078-2489/12/3/121 |
_version_ | 1797541599306055680 |
---|---|
author | Sichen Li Mélissa Zacharias Jochem Snuverink Jaime Coello de Portugal Fernando Perez-Cruz Davide Reggiani Andreas Adelmann |
author_facet | Sichen Li Mélissa Zacharias Jochem Snuverink Jaime Coello de Portugal Fernando Perez-Cruz Davide Reggiani Andreas Adelmann |
author_sort | Sichen Li |
collection | DOAJ |
description | The beam interruptions (interlocks) of particle accelerators, despite being necessary safety measures, lead to abrupt operational changes and a substantial loss of beam time. A novel time series classification approach is applied to decrease beam time loss in the High-Intensity Proton Accelerator complex by forecasting interlock events. The forecasting is performed through binary classification of windows of multivariate time series. The time series are transformed into Recurrence Plots which are then classified by a Convolutional Neural Network, which not only captures the inner structure of the time series but also uses the advances of image classification techniques. Our best-performing interlock-to-stable classifier reaches an Area under the ROC Curve value of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>0.71</mn><mo>±</mo><mn>0.01</mn></mrow></semantics></math></inline-formula> compared to <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>0.65</mn><mo>±</mo><mn>0.01</mn></mrow></semantics></math></inline-formula> of a Random Forest model, and it can potentially reduce the beam time loss by <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>0.5</mn><mo>±</mo><mn>0.2</mn></mrow></semantics></math></inline-formula> s per interlock. |
first_indexed | 2024-03-10T13:18:15Z |
format | Article |
id | doaj.art-457e13c0b2a7457d9d59c9c4a105302e |
institution | Directory Open Access Journal |
issn | 2078-2489 |
language | English |
last_indexed | 2024-03-10T13:18:15Z |
publishDate | 2021-03-01 |
publisher | MDPI AG |
record_format | Article |
series | Information |
spelling | doaj.art-457e13c0b2a7457d9d59c9c4a105302e2023-11-21T10:13:02ZengMDPI AGInformation2078-24892021-03-0112312110.3390/info12030121A Novel Approach for Classification and Forecasting of Time Series in Particle AcceleratorsSichen Li0Mélissa Zacharias1Jochem Snuverink2Jaime Coello de Portugal3Fernando Perez-Cruz4Davide Reggiani5Andreas Adelmann6Paul Scherrer Institut, 5232 Villigen, SwitzerlandPaul Scherrer Institut, 5232 Villigen, SwitzerlandPaul Scherrer Institut, 5232 Villigen, SwitzerlandPaul Scherrer Institut, 5232 Villigen, SwitzerlandSwiss Data Science Center, ETH Zürich and EPFL, Universitätstrasse 25, 8092 Zürich, SwitzerlandPaul Scherrer Institut, 5232 Villigen, SwitzerlandPaul Scherrer Institut, 5232 Villigen, SwitzerlandThe beam interruptions (interlocks) of particle accelerators, despite being necessary safety measures, lead to abrupt operational changes and a substantial loss of beam time. A novel time series classification approach is applied to decrease beam time loss in the High-Intensity Proton Accelerator complex by forecasting interlock events. The forecasting is performed through binary classification of windows of multivariate time series. The time series are transformed into Recurrence Plots which are then classified by a Convolutional Neural Network, which not only captures the inner structure of the time series but also uses the advances of image classification techniques. Our best-performing interlock-to-stable classifier reaches an Area under the ROC Curve value of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>0.71</mn><mo>±</mo><mn>0.01</mn></mrow></semantics></math></inline-formula> compared to <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>0.65</mn><mo>±</mo><mn>0.01</mn></mrow></semantics></math></inline-formula> of a Random Forest model, and it can potentially reduce the beam time loss by <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>0.5</mn><mo>±</mo><mn>0.2</mn></mrow></semantics></math></inline-formula> s per interlock.https://www.mdpi.com/2078-2489/12/3/121time series classificationrecurrence plotconvolutional neural networkrandom forestcharged particle accelerator |
spellingShingle | Sichen Li Mélissa Zacharias Jochem Snuverink Jaime Coello de Portugal Fernando Perez-Cruz Davide Reggiani Andreas Adelmann A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators Information time series classification recurrence plot convolutional neural network random forest charged particle accelerator |
title | A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators |
title_full | A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators |
title_fullStr | A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators |
title_full_unstemmed | A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators |
title_short | A Novel Approach for Classification and Forecasting of Time Series in Particle Accelerators |
title_sort | novel approach for classification and forecasting of time series in particle accelerators |
topic | time series classification recurrence plot convolutional neural network random forest charged particle accelerator |
url | https://www.mdpi.com/2078-2489/12/3/121 |
work_keys_str_mv | AT sichenli anovelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT melissazacharias anovelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT jochemsnuverink anovelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT jaimecoellodeportugal anovelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT fernandoperezcruz anovelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT davidereggiani anovelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT andreasadelmann anovelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT sichenli novelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT melissazacharias novelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT jochemsnuverink novelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT jaimecoellodeportugal novelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT fernandoperezcruz novelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT davidereggiani novelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators AT andreasadelmann novelapproachforclassificationandforecastingoftimeseriesinparticleaccelerators |