A minimum relative entropy-based density selection scheme for bayesian estimations of energy-related problems

Xiaoran Shi, Nurcin Celik

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

3 Citations (Scopus)

Abstract

Particle filtering procedures are widely used in Bayesian state estimation of large-scale dynamic systems given their massive datasets. However, they may end up in a situation called degeneracy, where a single particle abruptly possesses significant amount of normalized weights. An efficacious selection of the importance sampling density plays a pivotal role in achieving good performance of the particle filters by preventing the sampling procedure from generating degenerated weights for particles at early stages of the process. Such a selection of a proposal distribution that takes into account both the transition prior and the likelihood becomes especially crucial for better estimation accuracy when the observation data has significant impacts on the posterior states. In this study, we propose a novel importance density selection structure for particle filters based on the minimized relative entropy principle. First, the theoretical derivation of the proposed structure is presented. Then, the proposed scheme is benchmarked against other sampling schemes that exist in the literature for state estimation in terms of their estimation qualities and computational efficiencies through a set of synthetic experiments. Finally, various energy-related estimation problems that match the applicable circumstances of the proposed scheme are briefly discussed.

Original languageEnglish
Title of host publication62nd IIE Annual Conference and Expo 2012
PublisherInstitute of Industrial Engineers
Pages769-778
Number of pages10
StatePublished - Jan 1 2012
Event62nd IIE Annual Conference and Expo 2012 - Orlando, FL, United States
Duration: May 19 2012May 23 2012

Other

Other62nd IIE Annual Conference and Expo 2012
CountryUnited States
CityOrlando, FL
Period5/19/125/23/12

Fingerprint

Entropy
State estimation
Sampling
Importance sampling
Computational efficiency
Dynamical systems
Experiments

Keywords

  • Importance density selection
  • Importance sampling
  • Relative entropy
  • Sequential monte carlo methods

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering

Cite this

Shi, X., & Celik, N. (2012). A minimum relative entropy-based density selection scheme for bayesian estimations of energy-related problems. In 62nd IIE Annual Conference and Expo 2012 (pp. 769-778). Institute of Industrial Engineers.

A minimum relative entropy-based density selection scheme for bayesian estimations of energy-related problems. / Shi, Xiaoran; Celik, Nurcin.

62nd IIE Annual Conference and Expo 2012. Institute of Industrial Engineers, 2012. p. 769-778.

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

Shi, X & Celik, N 2012, A minimum relative entropy-based density selection scheme for bayesian estimations of energy-related problems. in 62nd IIE Annual Conference and Expo 2012. Institute of Industrial Engineers, pp. 769-778, 62nd IIE Annual Conference and Expo 2012, Orlando, FL, United States, 5/19/12.
Shi X, Celik N. A minimum relative entropy-based density selection scheme for bayesian estimations of energy-related problems. In 62nd IIE Annual Conference and Expo 2012. Institute of Industrial Engineers. 2012. p. 769-778
Shi, Xiaoran ; Celik, Nurcin. / A minimum relative entropy-based density selection scheme for bayesian estimations of energy-related problems. 62nd IIE Annual Conference and Expo 2012. Institute of Industrial Engineers, 2012. pp. 769-778
@inproceedings{719c17329dc44b24ba79d93f72dd11ca,
title = "A minimum relative entropy-based density selection scheme for bayesian estimations of energy-related problems",
abstract = "Particle filtering procedures are widely used in Bayesian state estimation of large-scale dynamic systems given their massive datasets. However, they may end up in a situation called degeneracy, where a single particle abruptly possesses significant amount of normalized weights. An efficacious selection of the importance sampling density plays a pivotal role in achieving good performance of the particle filters by preventing the sampling procedure from generating degenerated weights for particles at early stages of the process. Such a selection of a proposal distribution that takes into account both the transition prior and the likelihood becomes especially crucial for better estimation accuracy when the observation data has significant impacts on the posterior states. In this study, we propose a novel importance density selection structure for particle filters based on the minimized relative entropy principle. First, the theoretical derivation of the proposed structure is presented. Then, the proposed scheme is benchmarked against other sampling schemes that exist in the literature for state estimation in terms of their estimation qualities and computational efficiencies through a set of synthetic experiments. Finally, various energy-related estimation problems that match the applicable circumstances of the proposed scheme are briefly discussed.",
keywords = "Importance density selection, Importance sampling, Relative entropy, Sequential monte carlo methods",
author = "Xiaoran Shi and Nurcin Celik",
year = "2012",
month = "1",
day = "1",
language = "English",
pages = "769--778",
booktitle = "62nd IIE Annual Conference and Expo 2012",
publisher = "Institute of Industrial Engineers",

}

TY - GEN

T1 - A minimum relative entropy-based density selection scheme for bayesian estimations of energy-related problems

AU - Shi, Xiaoran

AU - Celik, Nurcin

PY - 2012/1/1

Y1 - 2012/1/1

N2 - Particle filtering procedures are widely used in Bayesian state estimation of large-scale dynamic systems given their massive datasets. However, they may end up in a situation called degeneracy, where a single particle abruptly possesses significant amount of normalized weights. An efficacious selection of the importance sampling density plays a pivotal role in achieving good performance of the particle filters by preventing the sampling procedure from generating degenerated weights for particles at early stages of the process. Such a selection of a proposal distribution that takes into account both the transition prior and the likelihood becomes especially crucial for better estimation accuracy when the observation data has significant impacts on the posterior states. In this study, we propose a novel importance density selection structure for particle filters based on the minimized relative entropy principle. First, the theoretical derivation of the proposed structure is presented. Then, the proposed scheme is benchmarked against other sampling schemes that exist in the literature for state estimation in terms of their estimation qualities and computational efficiencies through a set of synthetic experiments. Finally, various energy-related estimation problems that match the applicable circumstances of the proposed scheme are briefly discussed.

AB - Particle filtering procedures are widely used in Bayesian state estimation of large-scale dynamic systems given their massive datasets. However, they may end up in a situation called degeneracy, where a single particle abruptly possesses significant amount of normalized weights. An efficacious selection of the importance sampling density plays a pivotal role in achieving good performance of the particle filters by preventing the sampling procedure from generating degenerated weights for particles at early stages of the process. Such a selection of a proposal distribution that takes into account both the transition prior and the likelihood becomes especially crucial for better estimation accuracy when the observation data has significant impacts on the posterior states. In this study, we propose a novel importance density selection structure for particle filters based on the minimized relative entropy principle. First, the theoretical derivation of the proposed structure is presented. Then, the proposed scheme is benchmarked against other sampling schemes that exist in the literature for state estimation in terms of their estimation qualities and computational efficiencies through a set of synthetic experiments. Finally, various energy-related estimation problems that match the applicable circumstances of the proposed scheme are briefly discussed.

KW - Importance density selection

KW - Importance sampling

KW - Relative entropy

KW - Sequential monte carlo methods

UR - http://www.scopus.com/inward/record.url?scp=84900339477&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84900339477&partnerID=8YFLogxK

M3 - Conference contribution

SP - 769

EP - 778

BT - 62nd IIE Annual Conference and Expo 2012

PB - Institute of Industrial Engineers

ER -