Comparison of Different Features and Neural Networks for Predicting Industrial Paper Press Condition

João Antunes Rodrigues; José Torres Farinha; Mateus Mendes; Ricardo J.G. Mateus; António J.Marques Cardoso

doi:10.3390/en15176308

Comparison of Different Features and Neural Networks for Predicting Industrial Paper Press Condition

João Antunes Rodrigues, José Torres Farinha, Mateus Mendes, Ricardo J.G. Mateus, António J.Marques Cardoso

EIGES - Research Center for Industrial Engineering, Management and Sustainability

Research output: Contribution to journal › Article › peer-review

9 Citations (Scopus)

Abstract

Forecasting has extreme importance in industry due to the numerous competitive advantages that it provides, allowing to foresee what might happen and adjust management decisions accordingly. Industries increasingly use sensors, which allow for large-scale data collection. Big datasets enable training, testing and application of complex predictive algorithms based on machine learning models. The present paper focuses on predicting values from sensors installed on a pulp paper press, using data collected over three years. The variables analyzed are electric current, pressure, temperature, torque, oil level and velocity. The results of XGBoost and artificial neural networks, with different feature vectors, are compared. They show that it is possible to predict sensor data in the long term and thus predict the asset’s behaviour several days in advance.

Original language	English
Article number	6308
Journal	Energies
Volume	15
Issue number	17
DOIs	https://doi.org/10.3390/en15176308
Publication status	Published - Sept 2022

Bibliographical note

Publisher Copyright:
© 2022 by the authors.

Funding

The research leading to these results has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowvska-Curie grant agreement 871284 project SSHARE and the European Regional Development Fund (ERDF) through the Operational Programme for Competitiveness and Internationalization (COMPETE 2020), under ProjectPOCI-01-0145-FEDER-029494, and by National Funds through the FCT—Portuguese Foundationfor Science and Technology, under Projects PTDC/EEI-EEE/29494/2017, UIDB/04131/2020, and UIDP/04131/2020. This research is sponsored by FEDER funds through the program COMPETE—Programa Operacional Factores de Competitividade—and by national funds through FCT—Fundação para a Ciência e a Tecnologia—under the project UIDB/00285/2020. This work was produced with the support of INCD funded by FCT and FEDER under the project 01/SAICT/2016 nº 022153.

Funders	Funder number
INCD	01/SAICT/2016, 022153
Horizon 2020 Framework Programme	871284
Fundação para a Ciência e a Tecnologia	UIDB/04131/2020, UIDB/00285/2020, PTDC/EEI-EEE/29494/2017, UIDP/04131/2020
European Regional Development Fund	ProjectPOCI-01-0145-FEDER-029494

Keywords

XGBoost
forecast
maintenance
neural networks
sensor prediction

Access to Document

10.3390/en15176308

Cite this

@article{a0f407cca1a240d9a5a330f690a90135,

title = "Comparison of Different Features and Neural Networks for Predicting Industrial Paper Press Condition",

abstract = "Forecasting has extreme importance in industry due to the numerous competitive advantages that it provides, allowing to foresee what might happen and adjust management decisions accordingly. Industries increasingly use sensors, which allow for large-scale data collection. Big datasets enable training, testing and application of complex predictive algorithms based on machine learning models. The present paper focuses on predicting values from sensors installed on a pulp paper press, using data collected over three years. The variables analyzed are electric current, pressure, temperature, torque, oil level and velocity. The results of XGBoost and artificial neural networks, with different feature vectors, are compared. They show that it is possible to predict sensor data in the long term and thus predict the asset{\textquoteright}s behaviour several days in advance.",

keywords = "XGBoost, forecast, maintenance, neural networks, sensor prediction",

author = "Rodrigues, {Jo{\~a}o Antunes} and Farinha, {Jos{\'e} Torres} and Mateus Mendes and Mateus, {Ricardo J.G.} and Cardoso, {Ant{\'o}nio J.Marques}",

note = "Publisher Copyright: {\textcopyright} 2022 by the authors.",

year = "2022",

month = sep,

doi = "10.3390/en15176308",

language = "English",

volume = "15",

journal = "Energies",

issn = "1996-1073",

publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",

number = "17",

}

TY - JOUR

T1 - Comparison of Different Features and Neural Networks for Predicting Industrial Paper Press Condition

AU - Rodrigues, João Antunes

AU - Farinha, José Torres

AU - Mendes, Mateus

AU - Mateus, Ricardo J.G.

AU - Cardoso, António J.Marques

PY - 2022/9

Y1 - 2022/9

N2 - Forecasting has extreme importance in industry due to the numerous competitive advantages that it provides, allowing to foresee what might happen and adjust management decisions accordingly. Industries increasingly use sensors, which allow for large-scale data collection. Big datasets enable training, testing and application of complex predictive algorithms based on machine learning models. The present paper focuses on predicting values from sensors installed on a pulp paper press, using data collected over three years. The variables analyzed are electric current, pressure, temperature, torque, oil level and velocity. The results of XGBoost and artificial neural networks, with different feature vectors, are compared. They show that it is possible to predict sensor data in the long term and thus predict the asset’s behaviour several days in advance.

AB - Forecasting has extreme importance in industry due to the numerous competitive advantages that it provides, allowing to foresee what might happen and adjust management decisions accordingly. Industries increasingly use sensors, which allow for large-scale data collection. Big datasets enable training, testing and application of complex predictive algorithms based on machine learning models. The present paper focuses on predicting values from sensors installed on a pulp paper press, using data collected over three years. The variables analyzed are electric current, pressure, temperature, torque, oil level and velocity. The results of XGBoost and artificial neural networks, with different feature vectors, are compared. They show that it is possible to predict sensor data in the long term and thus predict the asset’s behaviour several days in advance.

KW - XGBoost

KW - forecast

KW - maintenance

KW - neural networks

KW - sensor prediction

UR - http://www.scopus.com/inward/record.url?scp=85138005842&partnerID=8YFLogxK

U2 - 10.3390/en15176308

DO - 10.3390/en15176308

M3 - Article

AN - SCOPUS:85138005842

SN - 1996-1073

VL - 15

JO - Energies

JF - Energies

IS - 17

M1 - 6308

ER -

Comparison of Different Features and Neural Networks for Predicting Industrial Paper Press Condition

Abstract

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this