Evaluation of filtrate water quality of a river bank filtration facility using neural computing techniques

Goloka Behari Sahoo, Chittaranjan Ray

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Riverbank filtration (RBF) is a natural and low-cost water-treatment process in which contaminants of surface water are removed or degraded as the river water moves through the adjoining alluvial aquifer into the pumping wells. Because river-aquifer interaction is highly nonlinear, time-varying, and spatially-distributed processes that is not easily described by simple mathematical models, application of artificial neural networks (ANNs) is examined to evaluate the effectiveness of a RBF facility at Louisville. Kentucky. USA. Three types of ANNs: feed-forward back-propagation network (BPN), radial basis function network (RBFN), and fuzzy inference system network (FISN) were used in this study. It is shown that BPN and RBFN predicted values were in excellent agreement with the measured values having correlation coefficient above 0.99, whereas FISN was able to predict only temperature and HPC removal of the filtrate water quality with correlation coefficient above 0.97.

Original languageEnglish (US)
Title of host publicationProceedings of the 2nd Indian International Conference on Artificial Intelligence, IICAI 2005
Pages970-986
Number of pages17
StatePublished - 2005
Externally publishedYes
Event2nd Indian International Conference on Artificial Intelligence, IICAI 2005 - Pune, India
Duration: Dec 20 2005Dec 22 2005

Publication series

NameProceedings of the 2nd Indian International Conference on Artificial Intelligence, IICAI 2005

Conference

Conference2nd Indian International Conference on Artificial Intelligence, IICAI 2005
Country/TerritoryIndia
CityPune
Period12/20/0512/22/05

ASJC Scopus subject areas

  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'Evaluation of filtrate water quality of a river bank filtration facility using neural computing techniques'. Together they form a unique fingerprint.

Cite this