Machine learning modelling of removal of reactive orange RO16 by chemical activated carbon in textile wastewater

Basit Öğe Kaydı
dc.authorid, Muhammad Bilal/0000-0003-3053-9149
dc.authoridKhan, Dr. Muhammad Saqib/0000-0002-0897-7436
dc.contributor.authorKhan, Izaz Ullah
dc.contributor.authorShah, Jehanzeb Ali
dc.contributor.authorBilal, Muhammad
dc.contributor.authorFaiza
dc.contributor.authorKhan, Muhammad Saqib
dc.contributor.authorShah, Sajid
dc.contributor.authorAkgul, Ali
dc.date.accessioned2024-12-24T19:31:00Z
dc.date.available2024-12-24T19:31:00Z
dc.date.issued2023
dc.departmentSiirt Üniversitesi
dc.description.abstractThis study develops machine learning model of removal of reactive orange dye (Azo) RO16 from textile wastewater by chemical activated carbon CAC. The study addresses the contamination removal efficiency with respect to changing dynamics of concentration, temperature, time, pH and dose, respectively. Machine learning based learning multiple polynomial regression is implemented to fit a model on the experimental observed data. The machine learns from the data and fit the multiple polynomial regression model for the data. The observed and predicted data are in close agreement with the R-squared value of 92%. The results show that the baseline efficiency of using chemical activated carbon adsorbent for removing RO16 is 76.5%. The most significant input parameter increasing the efficiency by a constant value of 35 units out of 100 is the second order response of the dose. Moreover, four input parameters can considerably increase the efficiency. Furthermore, six input parameters can considerably decrease the efficiency. It is investigated, that the second order response with respect to time has the minute decreasing effect on the removal efficiency. The superior abilities of the modeling are two fold. Firstly, the contamination removal of reactive orange dye (Azo) RO16 with chemical activated carbon adsorbent is studied with respect to five multiple parameters. Secondly, the model exploits the machine learning capability of the renowned Python machine learning module sklearn to fit a multiple polynomial regression model. Thus a robust model is fitted giving twenty-one inputs/output interactions and responses. From the input-target correlation analysis it is clear that the removal efficiency has a strong correlation with the time. It has considerably significant relationship with dose of the CAC and the temperature with values of 18% and 17%, respectively. Moreover, the removal efficiency has inverse relations with pH and Ci, with values of 15% and 12%, respectively.
dc.identifier.doi10.3233/JIFS-220781
dc.identifier.endpage7993
dc.identifier.issn1064-1246
dc.identifier.issn1875-8967
dc.identifier.issue5
dc.identifier.scopus2-s2.0-85166668534
dc.identifier.scopusqualityQ1
dc.identifier.startpage7977
dc.identifier.urihttps://doi.org/10.3233/JIFS-220781
dc.identifier.urihttps://hdl.handle.net/20.500.12604/7771
dc.identifier.volume44
dc.identifier.wosWOS:000980903000066
dc.identifier.wosqualityQ3
dc.indekslendigikaynakWeb of Science
dc.indekslendigikaynakScopus
dc.language.isoen
dc.publisherIos Press
dc.relation.ispartofJournal of Intelligent & Fuzzy Systems
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/closedAccess
dc.snmzKA_20241222
dc.subjectTextile wastewater
dc.subjectreactive orange dye
dc.subjectcontamination removal
dc.subjectchemical activated carbon
dc.subjectmachine learning
dc.subjectmultiple polynomial regression
dc.titleMachine learning modelling of removal of reactive orange RO16 by chemical activated carbon in textile wastewater
dc.typeArticle

Dosyalar

Koleksiyon

WOS İndeksli Yayınlar Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu