Failure prognosis using nonlinear short-time prediction and multi-time scale recursive estimation

David Chelidze, Joseph Paul Cusumano, Anindya Chatterjee

Research output: Contribution to conferencePaper

3 Citations (Scopus)

Abstract

In this paper we describe a new, general purpose machinery diagnostic/prognostic algorithm for tracking and predicting evolving damage using only available "macroscopic" observable quantities. The damage is viewed as occurring in a hierarchical dynamical system consisting of a directly observable, "fast" subsystem coupled with a hidden, "slow" subsystem describing damage evolution. This method provides damage diagnostics and failure prognostics requiring only the measurements from the fast subsystem and a model of the slow subsystem. Damage tracking is accomplished by a two-time-scale modeling strategy based on phase space reconstruction using the measured fast-time data. Short-time predictive models are constructed using the reconstructed phase space of the reference (undamaged) fast sub-system. Later, fast-time data for the damaged system is collected and used to estimate the short-time reference model prediction error, or a tracking function. An average value of the tracking function over a given data record is used as a tracking metric, or measure of the current damage state. Recursive, nonlinear filtering is used to estimate the actual damage state based on the tracking metric input. Estimates of remaining useful life are obtained recursively using a linear Kalman filter. This method is applied to an experimental nonlinear oscillator containing a beam with a crack which propagates to complete failure. We demonstrate the ability to track the evolving damage state of the beam using only strain time series data. We also give accurate predictions of remaining useful life, in real time, beginning well in advance of the final complete fracture at the end of experiment.

Original languageEnglish (US)
Pages901-910
Number of pages10
StatePublished - Dec 1 2001
Event18th Biennial Conference on Mechanical Vibration and Noise - Pittsburgh, PA, United States
Duration: Sep 9 2001Sep 12 2001

Other

Other18th Biennial Conference on Mechanical Vibration and Noise
CountryUnited States
CityPittsburgh, PA
Period9/9/019/12/01

Fingerprint

Recursive Estimation
Prognosis
Time Scales
Damage
Prediction
Subsystem
Nonlinear filtering
Kalman filters
Machinery
Time series
Dynamical systems
Cracks
Diagnostics
Estimate
Phase Space Reconstruction
Metric
Linear Filter
Nonlinear Filtering
Model Error
Nonlinear Oscillator

All Science Journal Classification (ASJC) codes

  • Modeling and Simulation
  • Mechanical Engineering
  • Computer Science Applications
  • Computer Graphics and Computer-Aided Design

Cite this

Chelidze, D., Cusumano, J. P., & Chatterjee, A. (2001). Failure prognosis using nonlinear short-time prediction and multi-time scale recursive estimation. 901-910. Paper presented at 18th Biennial Conference on Mechanical Vibration and Noise, Pittsburgh, PA, United States.
Chelidze, David ; Cusumano, Joseph Paul ; Chatterjee, Anindya. / Failure prognosis using nonlinear short-time prediction and multi-time scale recursive estimation. Paper presented at 18th Biennial Conference on Mechanical Vibration and Noise, Pittsburgh, PA, United States.10 p.
@conference{cec5893041ff4321b3cd38fe60efcf3c,
title = "Failure prognosis using nonlinear short-time prediction and multi-time scale recursive estimation",
abstract = "In this paper we describe a new, general purpose machinery diagnostic/prognostic algorithm for tracking and predicting evolving damage using only available {"}macroscopic{"} observable quantities. The damage is viewed as occurring in a hierarchical dynamical system consisting of a directly observable, {"}fast{"} subsystem coupled with a hidden, {"}slow{"} subsystem describing damage evolution. This method provides damage diagnostics and failure prognostics requiring only the measurements from the fast subsystem and a model of the slow subsystem. Damage tracking is accomplished by a two-time-scale modeling strategy based on phase space reconstruction using the measured fast-time data. Short-time predictive models are constructed using the reconstructed phase space of the reference (undamaged) fast sub-system. Later, fast-time data for the damaged system is collected and used to estimate the short-time reference model prediction error, or a tracking function. An average value of the tracking function over a given data record is used as a tracking metric, or measure of the current damage state. Recursive, nonlinear filtering is used to estimate the actual damage state based on the tracking metric input. Estimates of remaining useful life are obtained recursively using a linear Kalman filter. This method is applied to an experimental nonlinear oscillator containing a beam with a crack which propagates to complete failure. We demonstrate the ability to track the evolving damage state of the beam using only strain time series data. We also give accurate predictions of remaining useful life, in real time, beginning well in advance of the final complete fracture at the end of experiment.",
author = "David Chelidze and Cusumano, {Joseph Paul} and Anindya Chatterjee",
year = "2001",
month = "12",
day = "1",
language = "English (US)",
pages = "901--910",
note = "18th Biennial Conference on Mechanical Vibration and Noise ; Conference date: 09-09-2001 Through 12-09-2001",

}

Chelidze, D, Cusumano, JP & Chatterjee, A 2001, 'Failure prognosis using nonlinear short-time prediction and multi-time scale recursive estimation', Paper presented at 18th Biennial Conference on Mechanical Vibration and Noise, Pittsburgh, PA, United States, 9/9/01 - 9/12/01 pp. 901-910.

Failure prognosis using nonlinear short-time prediction and multi-time scale recursive estimation. / Chelidze, David; Cusumano, Joseph Paul; Chatterjee, Anindya.

2001. 901-910 Paper presented at 18th Biennial Conference on Mechanical Vibration and Noise, Pittsburgh, PA, United States.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Failure prognosis using nonlinear short-time prediction and multi-time scale recursive estimation

AU - Chelidze, David

AU - Cusumano, Joseph Paul

AU - Chatterjee, Anindya

PY - 2001/12/1

Y1 - 2001/12/1

N2 - In this paper we describe a new, general purpose machinery diagnostic/prognostic algorithm for tracking and predicting evolving damage using only available "macroscopic" observable quantities. The damage is viewed as occurring in a hierarchical dynamical system consisting of a directly observable, "fast" subsystem coupled with a hidden, "slow" subsystem describing damage evolution. This method provides damage diagnostics and failure prognostics requiring only the measurements from the fast subsystem and a model of the slow subsystem. Damage tracking is accomplished by a two-time-scale modeling strategy based on phase space reconstruction using the measured fast-time data. Short-time predictive models are constructed using the reconstructed phase space of the reference (undamaged) fast sub-system. Later, fast-time data for the damaged system is collected and used to estimate the short-time reference model prediction error, or a tracking function. An average value of the tracking function over a given data record is used as a tracking metric, or measure of the current damage state. Recursive, nonlinear filtering is used to estimate the actual damage state based on the tracking metric input. Estimates of remaining useful life are obtained recursively using a linear Kalman filter. This method is applied to an experimental nonlinear oscillator containing a beam with a crack which propagates to complete failure. We demonstrate the ability to track the evolving damage state of the beam using only strain time series data. We also give accurate predictions of remaining useful life, in real time, beginning well in advance of the final complete fracture at the end of experiment.

AB - In this paper we describe a new, general purpose machinery diagnostic/prognostic algorithm for tracking and predicting evolving damage using only available "macroscopic" observable quantities. The damage is viewed as occurring in a hierarchical dynamical system consisting of a directly observable, "fast" subsystem coupled with a hidden, "slow" subsystem describing damage evolution. This method provides damage diagnostics and failure prognostics requiring only the measurements from the fast subsystem and a model of the slow subsystem. Damage tracking is accomplished by a two-time-scale modeling strategy based on phase space reconstruction using the measured fast-time data. Short-time predictive models are constructed using the reconstructed phase space of the reference (undamaged) fast sub-system. Later, fast-time data for the damaged system is collected and used to estimate the short-time reference model prediction error, or a tracking function. An average value of the tracking function over a given data record is used as a tracking metric, or measure of the current damage state. Recursive, nonlinear filtering is used to estimate the actual damage state based on the tracking metric input. Estimates of remaining useful life are obtained recursively using a linear Kalman filter. This method is applied to an experimental nonlinear oscillator containing a beam with a crack which propagates to complete failure. We demonstrate the ability to track the evolving damage state of the beam using only strain time series data. We also give accurate predictions of remaining useful life, in real time, beginning well in advance of the final complete fracture at the end of experiment.

UR - http://www.scopus.com/inward/record.url?scp=0035791685&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035791685&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0035791685

SP - 901

EP - 910

ER -

Chelidze D, Cusumano JP, Chatterjee A. Failure prognosis using nonlinear short-time prediction and multi-time scale recursive estimation. 2001. Paper presented at 18th Biennial Conference on Mechanical Vibration and Noise, Pittsburgh, PA, United States.