Active learning for linear parameter-varying system identification

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

Authors

  • Robert Chin
  • Alejandro I. Maass
  • Nalika Ulapane
  • Chris Manzie
  • Iman Shames
  • Dragan Nesic
  • Hayato Nakada

Colleges, School and Institutes

External organisations

  • University of Melbourne
  • Toyota Motor Corporation
  • The Alan Turing Institute

Abstract

Active learning is proposed for selection of the next operating points in the design of experiments, for identifying linear parameter-varying systems. We extend existing approaches found in literature to multiple-input multiple-output systems with a multivariate scheduling parameter. Our approach is based on exploiting the probabilistic features of Gaussian process regression to quantify the overall model uncertainty across locally identified models. This results in a flexible framework which accommodates for various techniques to be applied for estimation of local linear models and their corresponding uncertainty. We perform active learning in application to the identification of a diesel engine air-path model, and demonstrate that measures of model uncertainty can be successfully reduced using the proposed framework.

Details

Original languageEnglish
Title of host publication21st IFAC World Congress (IFAC-V 2020)
Publication statusAccepted/In press - 8 May 2020
Event21st IFAC World Congress (IFAC-V 2020) - Virtual Event
Duration: 12 Jul 202017 Jul 2020

Publication series

NameIFAC-PapersOnLine
PublisherElsevier
ISSN (Electronic)2405-8963

Conference

Conference21st IFAC World Congress (IFAC-V 2020)
CityVirtual Event
Period12/07/2017/07/20

Keywords

  • Machine learning, System identification, Parameter estimation, Uncertainty, Diesel engines