Estimating the parameters of dynamical systems from Big Data using Sequential Monte Carlo samplers



Green, PL and Maskell, S
(2017) Estimating the parameters of dynamical systems from Big Data using Sequential Monte Carlo samplers. Mechanical Systems and Signal Processing, 93. 379 - 396.

This is the latest version of this item.

[img] Text
MSSP_SMC_preprint.pdf - Submitted Version

Download (952kB)
[img] Text
MSSP_SMC.pdf - Accepted Version

Download (716kB)

Abstract

In this paper the authors present a method which facilitates computationally efficientparameter estimation of dynamical systems from a continuously growing set of measure-ment data. It is shown that the proposed method, which utilises Sequential Monte Carlosamplers, is guaranteed to be fully parallelisable (in contrast to Markov chain MonteCarlo methods) and can be applied to a wide variety of scenarios within structural dynam-ics. Its ability to allowconvergenceof one’s parameter estimates, as more data is analysed,sets it apart from other sequential methods (such as the particle filter).

Item Type: Article
Uncontrolled Keywords: Big Data, Parameter estimation, Model updating, System identification, Sequential Monte Carlo sampler
Depositing User: Symplectic Admin
Date Deposited: 18 Sep 2017 08:36
Last Modified: 30 Sep 2021 20:25
DOI: 10.1016/j.ymssp.2016.12.023
Related URLs:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3009426

Available Versions of this Item