A Hybrid State Particle Filter for Failure Prognosis in Deteriorating Systems

Erotokritos Skordilis, Ramin Moghaddass

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

Abstract

In this paper, we present a new approach for predictive analytics using time-series condition-monitoring data. A double hybrid state-space model was proposed, with a continuous latent state vector serving as a unit-less health indicator regarding the health status of a system and a discrete latent state vector denoting the operating condition of the system through time. We also considered two observation levels. The first level consisted of continuous noisy sensor observations. The survivability of the system was considered as a second level of observations. Instead of an explicit formula, a non-parametric method relating the continuous latent state to the sensor observations using a single-layer feed-forward neural network, called the Extreme Learning Machine (ELM), was considered and a new Bayesian iterative procedure using the Expectation-Maximization algorithm for ELM training and joint state-parameter estimation was presented. Finally, our method was tested on a number of simulated degradation data sets.

Original languageEnglish (US)
Title of host publication2018 Annual Reliability and Maintainability Symposium, RAMS 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Volume2018-January
ISBN (Print)9781538628706
DOIs
StatePublished - Sep 11 2018
Event2018 Annual Reliability and Maintainability Symposium, RAMS 2018 - Reno, United States
Duration: Jan 22 2018Jan 25 2018

Other

Other2018 Annual Reliability and Maintainability Symposium, RAMS 2018
CountryUnited States
CityReno
Period1/22/181/25/18

Fingerprint

Prognosis
Particle Filter
Learning systems
Health
Extreme Learning Machine
Feedforward neural networks
Sensors
Condition monitoring
Parameter estimation
Time series
Degradation
Sensor
Condition Monitoring
Survivability
Nonparametric Methods
Feedforward Neural Networks
Expectation-maximization Algorithm
Hybrid Model
State-space Model
State Estimation

Keywords

  • Extreme Learning Machine
  • Logistic Regression
  • Particle Filter
  • Reliability Analysis

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality
  • Mathematics(all)
  • Computer Science Applications

Cite this

Skordilis, E., & Moghaddass, R. (2018). A Hybrid State Particle Filter for Failure Prognosis in Deteriorating Systems. In 2018 Annual Reliability and Maintainability Symposium, RAMS 2018 (Vol. 2018-January). [8463033] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/RAM.2018.8463033

A Hybrid State Particle Filter for Failure Prognosis in Deteriorating Systems. / Skordilis, Erotokritos; Moghaddass, Ramin.

2018 Annual Reliability and Maintainability Symposium, RAMS 2018. Vol. 2018-January Institute of Electrical and Electronics Engineers Inc., 2018. 8463033.

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

Skordilis, E & Moghaddass, R 2018, A Hybrid State Particle Filter for Failure Prognosis in Deteriorating Systems. in 2018 Annual Reliability and Maintainability Symposium, RAMS 2018. vol. 2018-January, 8463033, Institute of Electrical and Electronics Engineers Inc., 2018 Annual Reliability and Maintainability Symposium, RAMS 2018, Reno, United States, 1/22/18. https://doi.org/10.1109/RAM.2018.8463033
Skordilis E, Moghaddass R. A Hybrid State Particle Filter for Failure Prognosis in Deteriorating Systems. In 2018 Annual Reliability and Maintainability Symposium, RAMS 2018. Vol. 2018-January. Institute of Electrical and Electronics Engineers Inc. 2018. 8463033 https://doi.org/10.1109/RAM.2018.8463033
Skordilis, Erotokritos ; Moghaddass, Ramin. / A Hybrid State Particle Filter for Failure Prognosis in Deteriorating Systems. 2018 Annual Reliability and Maintainability Symposium, RAMS 2018. Vol. 2018-January Institute of Electrical and Electronics Engineers Inc., 2018.
@inproceedings{d3e71e89cfbb452a9d54f57d9f0cc917,
title = "A Hybrid State Particle Filter for Failure Prognosis in Deteriorating Systems",
abstract = "In this paper, we present a new approach for predictive analytics using time-series condition-monitoring data. A double hybrid state-space model was proposed, with a continuous latent state vector serving as a unit-less health indicator regarding the health status of a system and a discrete latent state vector denoting the operating condition of the system through time. We also considered two observation levels. The first level consisted of continuous noisy sensor observations. The survivability of the system was considered as a second level of observations. Instead of an explicit formula, a non-parametric method relating the continuous latent state to the sensor observations using a single-layer feed-forward neural network, called the Extreme Learning Machine (ELM), was considered and a new Bayesian iterative procedure using the Expectation-Maximization algorithm for ELM training and joint state-parameter estimation was presented. Finally, our method was tested on a number of simulated degradation data sets.",
keywords = "Extreme Learning Machine, Logistic Regression, Particle Filter, Reliability Analysis",
author = "Erotokritos Skordilis and Ramin Moghaddass",
year = "2018",
month = "9",
day = "11",
doi = "10.1109/RAM.2018.8463033",
language = "English (US)",
isbn = "9781538628706",
volume = "2018-January",
booktitle = "2018 Annual Reliability and Maintainability Symposium, RAMS 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - A Hybrid State Particle Filter for Failure Prognosis in Deteriorating Systems

AU - Skordilis, Erotokritos

AU - Moghaddass, Ramin

PY - 2018/9/11

Y1 - 2018/9/11

N2 - In this paper, we present a new approach for predictive analytics using time-series condition-monitoring data. A double hybrid state-space model was proposed, with a continuous latent state vector serving as a unit-less health indicator regarding the health status of a system and a discrete latent state vector denoting the operating condition of the system through time. We also considered two observation levels. The first level consisted of continuous noisy sensor observations. The survivability of the system was considered as a second level of observations. Instead of an explicit formula, a non-parametric method relating the continuous latent state to the sensor observations using a single-layer feed-forward neural network, called the Extreme Learning Machine (ELM), was considered and a new Bayesian iterative procedure using the Expectation-Maximization algorithm for ELM training and joint state-parameter estimation was presented. Finally, our method was tested on a number of simulated degradation data sets.

AB - In this paper, we present a new approach for predictive analytics using time-series condition-monitoring data. A double hybrid state-space model was proposed, with a continuous latent state vector serving as a unit-less health indicator regarding the health status of a system and a discrete latent state vector denoting the operating condition of the system through time. We also considered two observation levels. The first level consisted of continuous noisy sensor observations. The survivability of the system was considered as a second level of observations. Instead of an explicit formula, a non-parametric method relating the continuous latent state to the sensor observations using a single-layer feed-forward neural network, called the Extreme Learning Machine (ELM), was considered and a new Bayesian iterative procedure using the Expectation-Maximization algorithm for ELM training and joint state-parameter estimation was presented. Finally, our method was tested on a number of simulated degradation data sets.

KW - Extreme Learning Machine

KW - Logistic Regression

KW - Particle Filter

KW - Reliability Analysis

UR - http://www.scopus.com/inward/record.url?scp=85054128757&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054128757&partnerID=8YFLogxK

U2 - 10.1109/RAM.2018.8463033

DO - 10.1109/RAM.2018.8463033

M3 - Conference contribution

SN - 9781538628706

VL - 2018-January

BT - 2018 Annual Reliability and Maintainability Symposium, RAMS 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -