Reference : Comparison of supervised learning techniques for atmospheric pollutant monitoring in ...
Scientific journals : Other
Engineering, computing & technology : Multidisciplinary, general & others
Comparison of supervised learning techniques for atmospheric pollutant monitoring in a Kraft pulp mill
Sainlez, Matthieu mailto [Université de Liège - ULiège > > > Form.doct. sc. ingé. (chim. appl. - Bologne)]
Heyen, Georges [Université de Liège - ULiège > Département de chimie appliquée > LASSC (Labo d'analyse et synthèse des systèmes chimiques) >]
Journal of Computational & Applied Mathematics
Elsevier Science
Yes (verified by ORBi)
The Netherlands
[en] In this paper, supervised learning techniques are compared to predict nitro-
gen oxide (NOx) pollutant emission from the recovery boiler of a Kraft pulp
mill. Starting from a large database of raw process data related to a Kraft
recovery boiler, we consider a regression problem in which we are trying to
predict the value of a continuous variable. Generalization is done on the
worst case configuration possible to make sure the model is adequate: the
training period concerns stationary operations while test periods mainly fo-
cus on NOx emissions during transient operations. This comparison involves
neural network techniques (i.e., multilayer perceptron and NARX network),
tree-based methods and multiple linear regression. We illustrate the potential
of a dynamic neural approach compared to the others in this task.
Preprint submitted (status: in revision)

File(s) associated to this reference

Fulltext file(s):

Restricted access
JCAM-review-SAINLEZ.pdfPreprint submitted to JCAM (status: in revision)Author preprint2.15 MBRequest copy

Bookmark and Share SFX Query

All documents in ORBi are protected by a user license.