A gaussian process based prognostics framework for composite structures

Yingtao Liu, Subhasish Mohanty, Aditi Chattopadhyay

Research output: Chapter in Book/Report/Conference proceedingConference contribution

8 Citations (Scopus)

Abstract

Prognostic algorithms indicate the remaining useful life based on fault detection and diagnosis through condition monitoring framework. Due to the wide-spread applications of advanced composite materials in industry, the importance of prognosis on composite materials is being acknowledged by the research community. Prognosis has the potential to significantly enhance structural monitoring and maintenance planning. In this paper, a Gaussian process based prognostics framework is presented. Both off-line and on-line methods combined state estimation and life prediction of composite beam subject to fatigue loading. The framework consists of three main steps: 1) data acquisition, 2) feature extraction, 3) damage state prediction and remaining useful life estimation. Active piezoelectric and acoustic emission (AE) sensing techniques are applied to monitor the damage states. Wavelet transform is used to extract the piezoelectric sensing features. The number of counts from AE system was used as a feature. Piezoelectric or AE sensing features are used to build the input and output space of the Gaussian process. The future damage states and remaining useful life are predicted by Gaussian process based off-line and on-line algorithms. Accuracy of the Gaussian process based prognosis method is improved by including more training sets. Piezoelectric and AE features are also used for the state prediction. In the test cases presented, the piezoelectric features lead to better prognosis results. On-line prognosis is completed sequentially by combining experimental and predicted features. On-line damage state prediction and remaining useful life estimation shows good correlation with experimental data at later stages of fatigue life.

Original languageEnglish (US)
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
Volume7286
DOIs
StatePublished - 2009
EventModeling, Signal Processing, and Control for Smart Structures 2009 - San Diego, CA, United States
Duration: Mar 11 2009Mar 12 2009

Other

OtherModeling, Signal Processing, and Control for Smart Structures 2009
CountryUnited States
CitySan Diego, CA
Period3/11/093/12/09

Fingerprint

prognosis
Composite Structures
Prognosis
composite structures
Acoustic emissions
Composite structures
Gaussian Process
Acoustic Emission
acoustic emission
Damage
damage
Sensing
Composite materials
predictions
Composite Materials
Fatigue of materials
composite materials
Prediction
Fault Detection and Diagnosis
Condition monitoring

Keywords

  • Feature extraction
  • Prognostics
  • Remaining useful life estimation
  • Wavelet transform

ASJC Scopus subject areas

  • Applied Mathematics
  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics

Cite this

Liu, Y., Mohanty, S., & Chattopadhyay, A. (2009). A gaussian process based prognostics framework for composite structures. In Proceedings of SPIE - The International Society for Optical Engineering (Vol. 7286). [72860J] https://doi.org/10.1117/12.815889

A gaussian process based prognostics framework for composite structures. / Liu, Yingtao; Mohanty, Subhasish; Chattopadhyay, Aditi.

Proceedings of SPIE - The International Society for Optical Engineering. Vol. 7286 2009. 72860J.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Liu, Y, Mohanty, S & Chattopadhyay, A 2009, A gaussian process based prognostics framework for composite structures. in Proceedings of SPIE - The International Society for Optical Engineering. vol. 7286, 72860J, Modeling, Signal Processing, and Control for Smart Structures 2009, San Diego, CA, United States, 3/11/09. https://doi.org/10.1117/12.815889
Liu Y, Mohanty S, Chattopadhyay A. A gaussian process based prognostics framework for composite structures. In Proceedings of SPIE - The International Society for Optical Engineering. Vol. 7286. 2009. 72860J https://doi.org/10.1117/12.815889
Liu, Yingtao ; Mohanty, Subhasish ; Chattopadhyay, Aditi. / A gaussian process based prognostics framework for composite structures. Proceedings of SPIE - The International Society for Optical Engineering. Vol. 7286 2009.
@inproceedings{48629121f6824c8eabc8decaece69909,
title = "A gaussian process based prognostics framework for composite structures",
abstract = "Prognostic algorithms indicate the remaining useful life based on fault detection and diagnosis through condition monitoring framework. Due to the wide-spread applications of advanced composite materials in industry, the importance of prognosis on composite materials is being acknowledged by the research community. Prognosis has the potential to significantly enhance structural monitoring and maintenance planning. In this paper, a Gaussian process based prognostics framework is presented. Both off-line and on-line methods combined state estimation and life prediction of composite beam subject to fatigue loading. The framework consists of three main steps: 1) data acquisition, 2) feature extraction, 3) damage state prediction and remaining useful life estimation. Active piezoelectric and acoustic emission (AE) sensing techniques are applied to monitor the damage states. Wavelet transform is used to extract the piezoelectric sensing features. The number of counts from AE system was used as a feature. Piezoelectric or AE sensing features are used to build the input and output space of the Gaussian process. The future damage states and remaining useful life are predicted by Gaussian process based off-line and on-line algorithms. Accuracy of the Gaussian process based prognosis method is improved by including more training sets. Piezoelectric and AE features are also used for the state prediction. In the test cases presented, the piezoelectric features lead to better prognosis results. On-line prognosis is completed sequentially by combining experimental and predicted features. On-line damage state prediction and remaining useful life estimation shows good correlation with experimental data at later stages of fatigue life.",
keywords = "Feature extraction, Prognostics, Remaining useful life estimation, Wavelet transform",
author = "Yingtao Liu and Subhasish Mohanty and Aditi Chattopadhyay",
year = "2009",
doi = "10.1117/12.815889",
language = "English (US)",
isbn = "9780819475466",
volume = "7286",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

}

TY - GEN

T1 - A gaussian process based prognostics framework for composite structures

AU - Liu, Yingtao

AU - Mohanty, Subhasish

AU - Chattopadhyay, Aditi

PY - 2009

Y1 - 2009

N2 - Prognostic algorithms indicate the remaining useful life based on fault detection and diagnosis through condition monitoring framework. Due to the wide-spread applications of advanced composite materials in industry, the importance of prognosis on composite materials is being acknowledged by the research community. Prognosis has the potential to significantly enhance structural monitoring and maintenance planning. In this paper, a Gaussian process based prognostics framework is presented. Both off-line and on-line methods combined state estimation and life prediction of composite beam subject to fatigue loading. The framework consists of three main steps: 1) data acquisition, 2) feature extraction, 3) damage state prediction and remaining useful life estimation. Active piezoelectric and acoustic emission (AE) sensing techniques are applied to monitor the damage states. Wavelet transform is used to extract the piezoelectric sensing features. The number of counts from AE system was used as a feature. Piezoelectric or AE sensing features are used to build the input and output space of the Gaussian process. The future damage states and remaining useful life are predicted by Gaussian process based off-line and on-line algorithms. Accuracy of the Gaussian process based prognosis method is improved by including more training sets. Piezoelectric and AE features are also used for the state prediction. In the test cases presented, the piezoelectric features lead to better prognosis results. On-line prognosis is completed sequentially by combining experimental and predicted features. On-line damage state prediction and remaining useful life estimation shows good correlation with experimental data at later stages of fatigue life.

AB - Prognostic algorithms indicate the remaining useful life based on fault detection and diagnosis through condition monitoring framework. Due to the wide-spread applications of advanced composite materials in industry, the importance of prognosis on composite materials is being acknowledged by the research community. Prognosis has the potential to significantly enhance structural monitoring and maintenance planning. In this paper, a Gaussian process based prognostics framework is presented. Both off-line and on-line methods combined state estimation and life prediction of composite beam subject to fatigue loading. The framework consists of three main steps: 1) data acquisition, 2) feature extraction, 3) damage state prediction and remaining useful life estimation. Active piezoelectric and acoustic emission (AE) sensing techniques are applied to monitor the damage states. Wavelet transform is used to extract the piezoelectric sensing features. The number of counts from AE system was used as a feature. Piezoelectric or AE sensing features are used to build the input and output space of the Gaussian process. The future damage states and remaining useful life are predicted by Gaussian process based off-line and on-line algorithms. Accuracy of the Gaussian process based prognosis method is improved by including more training sets. Piezoelectric and AE features are also used for the state prediction. In the test cases presented, the piezoelectric features lead to better prognosis results. On-line prognosis is completed sequentially by combining experimental and predicted features. On-line damage state prediction and remaining useful life estimation shows good correlation with experimental data at later stages of fatigue life.

KW - Feature extraction

KW - Prognostics

KW - Remaining useful life estimation

KW - Wavelet transform

UR - http://www.scopus.com/inward/record.url?scp=69649098264&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=69649098264&partnerID=8YFLogxK

U2 - 10.1117/12.815889

DO - 10.1117/12.815889

M3 - Conference contribution

SN - 9780819475466

VL - 7286

BT - Proceedings of SPIE - The International Society for Optical Engineering

ER -