Bayesian system identification of dynamical systems using highly informative training data. (May 2015)

• Record Type: Journal Article
• Authors: Green, P.L.; Cross, E.J.; Worden, K.
• Is Part Of: Mechanical systems and signal processing. Volume 56/57(2015)
• Page Start: 109
• Page End: 122
• Publication Date: 2015-05
• Subjects: Nonlinear system identification -- Bayesian inference -- Markov chain Monte Carlo -- Shannon entropy -- Tamar bridge; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2014.10.003
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: This paper is concerned with the Bayesian system identification of structural dynamical systems using experimentally obtained training data. It is motivated by situations where, from a large quantity of training data, one must select a subset to infer probabilistic models. To that end, using concepts from information theory, expressions are derived which allow one to approximate the effect that a set of training data will have on parameter uncertainty as well as the plausibility of candidate model structures. The usefulness of this concept is then demonstrated through the system identification of several dynamical systems using both physics-based and emulator models. The result is a rigorous scientific framework which can be used to select 'highly informative' subsets from large quantities of training data. Abstract : Highlights: This paper addresses the situation where, with the aim of conducting Bayesian system identification, one is presented with very large sets of training data. Techniques are developed which allow one to choose a subset of the available training data which is 'highly informative' with regards to both levels of Bayesian inference – parameter estimation and model selection. Examples include the system identification of various systems using both physics-based and data-based models.