Variational system identification for nonlinear state-space models

DOI10.1016/J.AUTOMATICA.2022.110687zbMATH Open1505.93044arXiv2012.05072OpenAlexW3171715856MaRDI QIDQ2103663FDOQ2103663

Authors: Jarrad Courts, Thomas B. Schön, Brett Ninness, Adrian G. Wills 

Publication date: 9 December 2022

Published in: Automatica (Search for Journal in Brave)

Abstract: This paper considers parameter estimation for nonlinear state-space models, which is an important but challenging problem. We address this challenge by employing a variational inference (VI) approach, which is a principled method that has deep connections to maximum likelihood estimation. This VI approach ultimately provides estimates of the model as solutions to an optimisation problem, which is deterministic, tractable and can be solved using standard optimisation tools. A specialisation of this approach for systems with additive Gaussian noise is also detailed. The proposed method is examined numerically on a range of simulated and real examples focusing on the robustness to parameter initialisation; additionally, favourable comparisons are performed against state-of-the-art alternatives.

Full work available at URL: https://arxiv.org/abs/2012.05072

Recommendations

• System identification of nonlinear state-space models
• System identification of linear parameter varying state-space models
• scientific article; zbMATH DE number 1836472
• scientific article; zbMATH DE number 3937075
• Further results on ``System identification of nonlinear state-space models
• Identification of Nonlinear State-Space Systems From Heterogeneous Datasets
• scientific article
• scientific article
• Identification of non-linear stochastic systems by state dependent parameter estimation
• Identification of nonlinear systems using polynomial nonlinear state space models

zbMATH Keywords

parameter estimationvariational inferencenonlinear modelssystem identificationassumed density

Mathematics Subject Classification ID

System identification (93B30) Nonlinear systems in control theory (93C10)

Cites Work

• Title not available (Why is that?)
• System identification of nonlinear state-space models
• Title not available (Why is that?)
• Title not available (Why is that?)
• On Information and Sufficiency
• Inference in hidden Markov models.
• Filtering via Simulation: Auxiliary Particle Filters
• Convergence of a stochastic approximation version of the EM algorithm
• Monte Carlo Smoothing for Nonlinear Time Series
• An introduction to variational methods for graphical models
• Title not available (Why is that?)
• High-degree cubature Kalman filter
• Identification of Hammerstein-Wiener models
• Identification of dynamical systems. An introduction with applications.
• Maximum likelihood and prediction error methods
• On convexification of system identification criteria
• Bayesian filtering and smoothing
• Frequency domain identification of continuous-time output error models. I: Uniformly sampled data and frequency function approximation
• On the smoothness of nonlinear system identification
• Regularized LTI system identification in the presence of outliers: a variational EM approach
• Stochastic quasi-Newton with line-search regularisation

Cited In (2)

• Title not available (Why is that?)
• Sparse Bayesian Nonlinear System Identification Using Variational Inference

Uses Software

• KNITRO
• Matlab
• System Identification Toolbox
• Optimization Toolbox

This page was built for publication: Variational system identification for nonlinear state-space models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2103663)