학술논문
A Novel Filtering Based Recursive Estimation Algorithm for Box-Jenkins Systems
Document Type
Article
Author
Source
(2022): 3903-3913.
Subject
Language
Korean
ISSN
15986446
Abstract
This paper presents a novel filtering-based multi-innovation estimation algorithm for output-error autoregressive moving average (i.e., Box-Jenkins) systems. The key is to filter the input-output data by means of a linear filter and to derive a filtered identification model by eliminating the correlated noise. The new filtered model is an output-error moving average model whose parameters are estimated by using the multi-innovation identification theory. The parameter estimates of the original system are finally computed by the estimates of the filtered model. Furthermore, the proposed algorithm is extended to identify multivariable Box-Jenkins systems. The simulation results show that the proposed algorithm has higher estimation accuracy than the auxiliary model-based multi-innovation gradient algorithm.
This paper presents a novel filtering-based multi-innovation estimation algorithm for output-error autoregressive moving average (i.e., Box-Jenkins) systems. The key is to filter the input-output data by means of a linear filter and to derive a filtered identification model by eliminating the correlated noise. The new filtered model is an output-error moving average model whose parameters are estimated by using the multi-innovation identification theory. The parameter estimates of the original system are finally computed by the estimates of the filtered model. Furthermore, the proposed algorithm is extended to identify multivariable Box-Jenkins systems. The simulation results show that the proposed algorithm has higher estimation accuracy than the auxiliary model-based multi-innovation gradient algorithm.
This paper presents a novel filtering-based multi-innovation estimation algorithm for output-error autoregressive moving average (i.e., Box-Jenkins) systems. The key is to filter the input-output data by means of a linear filter and to derive a filtered identification model by eliminating the correlated noise. The new filtered model is an output-error moving average model whose parameters are estimated by using the multi-innovation identification theory. The parameter estimates of the original system are finally computed by the estimates of the filtered model. Furthermore, the proposed algorithm is extended to identify multivariable Box-Jenkins systems. The simulation results show that the proposed algorithm has higher estimation accuracy than the auxiliary model-based multi-innovation gradient algorithm.