Posterior density estimation for structural parameters using improved differential evolution adaptive Metropolis algorithm

Jin Zhou, Akira Mita, Liu Mei

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

The major difficulty of using Bayesian probabilistic inference for system identification is to obtain the posterior probability density of parameters conditioned by the measured response. The posterior density of structural parameters indicates how plausible each model is when considering the uncertainty of prediction errors. The Markov chain Monte Carlo (MCMC) method is a widespread medium for posterior inference but its convergence is often slow. The differential evolution adaptive Metropolis-Hasting (DREAM) algorithm boasts a population-based mechanism, which runs multiple different Markov chains simultaneously, and a global optimum exploration ability. This paper proposes an improved differential evolution adaptive Metropolis-Hasting algorithm (IDREAM) strategy to estimate the posterior density of structural parameters. The main benefit of IDREAM is its efficient MCMC simulation through its use of the adaptive Metropolis (AM) method with a mutation strategy for ensuring quick convergence and robust solutions. Its effectiveness was demonstrated in simulations on identifying the structural parameters with limited output data and noise polluted measurements.

Original languageEnglish
Pages (from-to)735-749
Number of pages15
JournalSmart Structures and Systems
Volume15
Issue number3
DOIs
Publication statusPublished - 2015 Mar 1

Fingerprint

Adaptive algorithms
Markov processes
Identification (control systems)
Monte Carlo methods

Keywords

  • Adaptive metropolis-hastings
  • Bayesian posterior probability density
  • Differential evolution
  • Markov chain Monte Carlo
  • Structural identification
  • Structural parameter estimation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Control and Systems Engineering
  • Computer Science Applications

Cite this

Posterior density estimation for structural parameters using improved differential evolution adaptive Metropolis algorithm. / Zhou, Jin; Mita, Akira; Mei, Liu.

In: Smart Structures and Systems, Vol. 15, No. 3, 01.03.2015, p. 735-749.

Research output: Contribution to journalArticle

@article{1d34d45607d14f5c83860078215e693d,
title = "Posterior density estimation for structural parameters using improved differential evolution adaptive Metropolis algorithm",
abstract = "The major difficulty of using Bayesian probabilistic inference for system identification is to obtain the posterior probability density of parameters conditioned by the measured response. The posterior density of structural parameters indicates how plausible each model is when considering the uncertainty of prediction errors. The Markov chain Monte Carlo (MCMC) method is a widespread medium for posterior inference but its convergence is often slow. The differential evolution adaptive Metropolis-Hasting (DREAM) algorithm boasts a population-based mechanism, which runs multiple different Markov chains simultaneously, and a global optimum exploration ability. This paper proposes an improved differential evolution adaptive Metropolis-Hasting algorithm (IDREAM) strategy to estimate the posterior density of structural parameters. The main benefit of IDREAM is its efficient MCMC simulation through its use of the adaptive Metropolis (AM) method with a mutation strategy for ensuring quick convergence and robust solutions. Its effectiveness was demonstrated in simulations on identifying the structural parameters with limited output data and noise polluted measurements.",
keywords = "Adaptive metropolis-hastings, Bayesian posterior probability density, Differential evolution, Markov chain Monte Carlo, Structural identification, Structural parameter estimation",
author = "Jin Zhou and Akira Mita and Liu Mei",
year = "2015",
month = "3",
day = "1",
doi = "10.12989/sss.2015.15.3.735",
language = "English",
volume = "15",
pages = "735--749",
journal = "Smart Structures and Systems",
issn = "1738-1584",
publisher = "Techno Press",
number = "3",

}

TY - JOUR

T1 - Posterior density estimation for structural parameters using improved differential evolution adaptive Metropolis algorithm

AU - Zhou, Jin

AU - Mita, Akira

AU - Mei, Liu

PY - 2015/3/1

Y1 - 2015/3/1

N2 - The major difficulty of using Bayesian probabilistic inference for system identification is to obtain the posterior probability density of parameters conditioned by the measured response. The posterior density of structural parameters indicates how plausible each model is when considering the uncertainty of prediction errors. The Markov chain Monte Carlo (MCMC) method is a widespread medium for posterior inference but its convergence is often slow. The differential evolution adaptive Metropolis-Hasting (DREAM) algorithm boasts a population-based mechanism, which runs multiple different Markov chains simultaneously, and a global optimum exploration ability. This paper proposes an improved differential evolution adaptive Metropolis-Hasting algorithm (IDREAM) strategy to estimate the posterior density of structural parameters. The main benefit of IDREAM is its efficient MCMC simulation through its use of the adaptive Metropolis (AM) method with a mutation strategy for ensuring quick convergence and robust solutions. Its effectiveness was demonstrated in simulations on identifying the structural parameters with limited output data and noise polluted measurements.

AB - The major difficulty of using Bayesian probabilistic inference for system identification is to obtain the posterior probability density of parameters conditioned by the measured response. The posterior density of structural parameters indicates how plausible each model is when considering the uncertainty of prediction errors. The Markov chain Monte Carlo (MCMC) method is a widespread medium for posterior inference but its convergence is often slow. The differential evolution adaptive Metropolis-Hasting (DREAM) algorithm boasts a population-based mechanism, which runs multiple different Markov chains simultaneously, and a global optimum exploration ability. This paper proposes an improved differential evolution adaptive Metropolis-Hasting algorithm (IDREAM) strategy to estimate the posterior density of structural parameters. The main benefit of IDREAM is its efficient MCMC simulation through its use of the adaptive Metropolis (AM) method with a mutation strategy for ensuring quick convergence and robust solutions. Its effectiveness was demonstrated in simulations on identifying the structural parameters with limited output data and noise polluted measurements.

KW - Adaptive metropolis-hastings

KW - Bayesian posterior probability density

KW - Differential evolution

KW - Markov chain Monte Carlo

KW - Structural identification

KW - Structural parameter estimation

UR - http://www.scopus.com/inward/record.url?scp=84925747317&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84925747317&partnerID=8YFLogxK

U2 - 10.12989/sss.2015.15.3.735

DO - 10.12989/sss.2015.15.3.735

M3 - Article

AN - SCOPUS:84925747317

VL - 15

SP - 735

EP - 749

JO - Smart Structures and Systems

JF - Smart Structures and Systems

SN - 1738-1584

IS - 3

ER -