Show simple item record

dc.contributor.authorWu, Di
dc.contributor.authorWang, Hao
dc.contributor.authorSeidu, Razak
dc.description.abstractA water supply system that integrates water source management, treatment and distribution is a critical infrastructure in urban areas. Traditional water quality research mostly focused on separate aspects, lacking a comprehensive coverage of all aspects, which undermines the prediction accuracies. In this paper, we propose a smart data analysis scheme to analyze and predict the water quality, considering all the water quality standard indicators. Instead of data output from water treatment, we collect the raw water data directly from water sources. We design two models to predict the water quality: (1) adaptive learning rate BP neural network (ALBP) and (2) 2-step isolation and random forest (2sIRF). We applied these models in the practical urban water supply systems of Oslo and Bergen in Norway. The results show that ALBP is theoretically simple and easy to implement. 2sIRF considers the risk distribution and shows higher prediction accuracy. In addition, we perform the correlation analysis of all the indicators and the importance analysis over different indicators. The domain experts have confirmed that this work is meaningful for future risk control and decision support in urban water supply systems.nb_NO
dc.rightsAttribution-NonCommercial-NoDerivatives 4.0 Internasjonal*
dc.titleSmart data driven quality prediction for urban water source managementnb_NO
dc.typeJournal articlenb_NO
dc.typePeer reviewednb_NO
dc.source.journalFuture generations computer systemsnb_NO
dc.description.localcode© 2020. This is the authors’ accepted and refereed manuscript to the article. Locked until 11.2.2022 due to copyright restrictions. This manuscript version is made available under the CC-BY-NC-ND 4.0 license
cristin.unitnameInstitutt for IKT og realfag
cristin.unitnameInstitutt for datateknologi og informatikk
cristin.unitnameInstitutt for havromsoperasjoner og byggteknikk

Files in this item


This item appears in the following Collection(s)

Show simple item record

Attribution-NonCommercial-NoDerivatives 4.0 Internasjonal
Except where otherwise noted, this item's license is described as Attribution-NonCommercial-NoDerivatives 4.0 Internasjonal