1990
DOI: 10.1002/cem.1180040203
|View full text |Cite
|
Sign up to set email alerts
|

Recursive parameter estimation

Abstract: SUMMARYThe use of recursive filtering techniques for parameter estimation in a variety of areas is reviewed. In particular, the Kalman filter algorithm is described, along with several variations, including square-root, UDUT and information filters. The solution to parameter estimation problems is discussed for both linear and non-linear models. Applications described include calibration, curve resolution in spectroscopy, chromatography, electrochemistry, kinetic analysis and process monitoring.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
3
0

Year Published

1992
1992
2015
2015

Publication Types

Select...
6
1

Relationship

0
7

Authors

Journals

citations
Cited by 17 publications
(3 citation statements)
references
References 67 publications
(33 reference statements)
0
3
0
Order By: Relevance
“…(8). H k in the measurement equation is given by x T k , where x k is the process measurement at time k. With the initial guess for the state being β c (β c is the regression coefficient vector of the current soft sensor model), the Kalman filter can perform a recursive estimate for the states (Rutan, 1990;Chen, 2003). A prior state estimate based on Eq.…”
Section: Mismatch Detection Of the Soft Sensor Model By Kalman Filtermentioning
confidence: 99%
See 1 more Smart Citation
“…(8). H k in the measurement equation is given by x T k , where x k is the process measurement at time k. With the initial guess for the state being β c (β c is the regression coefficient vector of the current soft sensor model), the Kalman filter can perform a recursive estimate for the states (Rutan, 1990;Chen, 2003). A prior state estimate based on Eq.…”
Section: Mismatch Detection Of the Soft Sensor Model By Kalman Filtermentioning
confidence: 99%
“…Nevertheless, choosing an appropriate forgetting factor for previous models is not a trivial task in recursive adaptation methods. Another alternative of performing online parameter estimation of soft sensor models is by utilizing Kalman filters, in which coefficient estimates are calculated by minimizing the noise effects (Rutan, 1990;Teppola et al, 1999). Some effort has been made to employ just-in-time learning (JITL) strategy to construct a local model based on a number of nearest neighbors of the test sample (query data) for adaptive predictions (Ge and Song, 2010).…”
Section: Introductionmentioning
confidence: 99%
“…Kalman filtering has found widespread use in analytical chemistry, e.g., for resolving overlapped responses, processing flow injection analysis (FIA) data and tackling drift problems, as demonstrated in several reviews on this topic. [16][17][18] In this work, the Kalman filtering algorithm was used to process RTP decay data and thus to resolve phosphorescent species whose RTP spectra are completely overlapped.…”
mentioning
confidence: 99%