<%pg=request.getParameter("page");%>
     

Franšais      print      e-mail    

Citation

Lamrini B., Le Lann M.-V., Lakhal E. K. and Benhammou A. (2007). Coagulation monitoring through a learning and expertise methodology for drinking water. v20 (4) : 325-338   [article in French]

Original Title: Dégradation photocatalytique du monolinuron et du linuron dans une suspension aqueuse de dioxide de titanium au contact de lumière solaire simulée.

Full text (PDF)

Abstracts

The present work proposes a learning classification method to identify the functional states of a coagulation process for the treatment of surface water and production of drinking water. Supervisory control and diagnosis were performed using the LAMDA (Learning Algorithm for Multivariate Data Analysis) classification technique. This expert learning method involves the processing and aggregation of all information stemming from an environmental process, and it allows the incorporation of the user’s knowledge. The study shows that it is possible to refine the diagnosis by taking into account the information obtained from common sensors (e.g., temperature, suspended solids, pH, conductivity, dissolved oxygen) together with the predicted coagulant dosage, as computed with an intelligent software sensor developed previously. The Rocade drinking water plant located at Marrakech, Morocco was chosen to test the method.

Keywords

Coagulation process, classification, supervised learning, unsupervised learning, pattern recognition, fuzzy logic.

Corresponding authors

Bouchra Lamrini, Laboratoire d’Automatique, de l’Environnement et de Procédés de Transfert (LAEPT) Faculté de Sciences Semlalia, BP: 2390, 40000-Marrakech, Maroc

Marie-Véronique Le Lann, LAAS-CNRS, 7-Avenue du Colonel Roche, 31077-Toulouse, cedex 4,
Université de Toulouse, INSA, DGEI, 135-Avenue de Rangueil, 31077-Toulouse Cedex 4, France
Courriel :
Courriel :

Franšais      print      e-mail    
     


Update: 2008-01-09
© INRS Eau, Terre et Environnement