Trend modelling with artificial neural networks. Case study: Operating zones identification for higher SO3 incorporation in cement clinker

Instantaneous measurements of process variables are usually not representative of the process effects as a whole when defining the condition of an output sample mainly in case of laboratory analysis. Moreover, process data have considerable dispersion. This leads to uncertainty in input–output time...

Full description

Saved in:
Bibliographic Details
Published inEngineering applications of artificial intelligence Vol. 54; pp. 17 - 25
Main Authors Lima, R.N., de Almeida, G.M., Braga, A.P., Cardoso, M.
Format Journal Article
LanguageEnglish
Published Elsevier Ltd 01.09.2016
Subjects
ANN
Instantaneous measurements
Process data dispersion
Time alignment
Trend modelling
Variable relationship
Variable relationship
ANN
Instantaneous measurements
Process data dispersion
Trend modelling
Time alignment
Online AccessGet full text

Cover

More Information
Summary:Instantaneous measurements of process variables are usually not representative of the process effects as a whole when defining the condition of an output sample mainly in case of laboratory analysis. Moreover, process data have considerable dispersion. This leads to uncertainty in input–output time alignment and in variable relationship. This work employs a trend data-based approach to overcome the negative effects of these uncertainties in both tasks variable selection commonly supported by correlation analysis and model identification. Two real case studies using a clinker rotary kiln from a cement plant and a chemical recovery boiler from a pulp mill were used for illustration purposes. More reliable data-driven system representation enhances the comprehension of the underlying system phenomena supporting a more rational basis for decision making.
ISSN:0952-1976
1873-6769
DOI:10.1016/j.engappai.2016.05.002