Modified Kalman filtering based multi-step-length gradient iterative algorithm for ARX models with random missing outputs

This study presents a modified Kalman filtering-based multi-step-length gradient iterative algorithm to identify ARX models with missing outputs. The Kalman filtering method is modified to enhance the estimation of unmeasurable outputs, laying the foundation for enabling the multi-step-length gradie...

Full description

Saved in:
Bibliographic Details
Published in:Automatica (Oxford) Vol. 118; p. 109034
Main Authors: Chen, Jing, Zhu, Quanmin, Liu, Yanjun
Format: Journal Article
Language:English
Published: Elsevier Ltd 01.08.2020
Subjects:
ARX model
Gradient iterative algorithm
Modified Kalman filter
Multi-step-length
Parameter estimation
Parameter estimation
Modified Kalman filter
ARX model
Multi-step-length
Gradient iterative algorithm
ISSN:0005-1098, 1873-2836
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This study presents a modified Kalman filtering-based multi-step-length gradient iterative algorithm to identify ARX models with missing outputs. The Kalman filtering method is modified to enhance the estimation of unmeasurable outputs, laying the foundation for enabling the multi-step-length gradient iterative algorithm to update effectively the ARX model parameter estimation through the estimated outputs. Compared to the classical gradient iterative algorithm, this study improves the estimation accuracy of the missing outputs by introducing a modified Kalman filter, and the parameter estimation convergence rate by deriving a new multi-step-length formulation. To validate the framework and the algorithm developed, a series of bench tests were conducted with computational experiments. The simulated numerical results are consistent with the analytically derived results in terms of the feasibility and effectiveness of the proposed procedure.
ISSN:0005-1098
1873-2836
DOI:10.1016/j.automatica.2020.109034