2015
DOI: 10.1016/j.snb.2014.09.113
|View full text |Cite
|
Sign up to set email alerts
|

Hierarchical methods to improve the performance of the SiC-FET as SO2 sensors in flue gas desulphurization systems

Abstract: Experiments were performed both in the laboratory and a desulfurization pilot unit in order to improve the SiC-FET sensor performance using two-step data evaluation. In both cases, a porous Pt-gate enhancement type SiC-FET was utilized in a temperature cycled operation (TCO). Liner Discriminant Analysis (LDA) was chosen as the method for multivariate data analysis. Hierarchical methods with two-step LDA worked quite well in the laboratory tests with SO 2 concentrations varied from 25-200 ppm. The same data eva… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
5
0

Year Published

2015
2015
2023
2023

Publication Types

Select...
5
2
1

Relationship

1
7

Authors

Journals

citations
Cited by 8 publications
(5 citation statements)
references
References 25 publications
0
5
0
Order By: Relevance
“…The quotient of the uncertainty divided by highest concentration is around 26 % for benzene and formaldehyde, and 20 % for naphthalene. Therefore we assume that the resolution of each model can be improved employing a hierarchical approach, similar to those used for classification with Linear Discriminant Analysis in [15] and [16]: first, the concentration is determined roughly with a wide-range model, and in a second step with a narrow-range model around this concentration. This approach is similar to local variants of the PLSR algorithm, like Locally Weighted PLSR (LW-PLSR) [17], which could therefore be another possibility to improve resolution (especially for naphthalene, as this also takes non-linearity into account).…”
Section: Resultsmentioning
confidence: 99%
“…The quotient of the uncertainty divided by highest concentration is around 26 % for benzene and formaldehyde, and 20 % for naphthalene. Therefore we assume that the resolution of each model can be improved employing a hierarchical approach, similar to those used for classification with Linear Discriminant Analysis in [15] and [16]: first, the concentration is determined roughly with a wide-range model, and in a second step with a narrow-range model around this concentration. This approach is similar to local variants of the PLSR algorithm, like Locally Weighted PLSR (LW-PLSR) [17], which could therefore be another possibility to improve resolution (especially for naphthalene, as this also takes non-linearity into account).…”
Section: Resultsmentioning
confidence: 99%
“…Besides, high temperature increases the catalytic processes efficiency on the electrode surface and thus worsens the sensors’ selectivity parameters. For example, in [ 39 ], to control SO 2 in emission desulfation systems in the energy sector, the authors propose to combat non-selectivity, i.e., reduce the O 2 , CO, and NO x influence by using the cyclic mode of the sensor heating up to 350–400 °C (to shift the maximum sensitivity in SO 2 favor) and carrying out linear discriminant analysis (LDA).…”
Section: Discussionmentioning
confidence: 99%
“…Especially in complex environments, it can be difficult to find one comprehensive model which works in all cases. It can, therefore, be beneficial to use several, more and more specialized, models to arrive at a prediction [56], [145], [180], [181]. For example, if humidity has a big influence, a first model could classify observations as low, medium, or high humidity, and subsequently feed the data to a specialized quantification model which 3.10.…”
Section: Hierarchical Modelsmentioning
confidence: 99%