skip to main content
Show Results with:

ANFIS-Based Prediction of the Decomposition of Sodium Aluminate Solutions in the Bayer Process

Savic, Marija et al.

Chemical engineering communications. Volume 203:Issue 8 (2016, August); pp 1053-1061 -- Taylor & Francis, Talylor & Francis Group

Online access

  • Title:
    ANFIS-Based Prediction of the Decomposition of Sodium Aluminate Solutions in the Bayer Process
  • Author: Savic, Marija;
    Mihajlovic, Ivan;
    Djordjevic, Predrag;
    Zivkovic, Zivan
  • Found In: Chemical engineering communications. Volume 203:Issue 8 (2016, August); pp 1053-1061
  • Journal Title: Chemical engineering communications
  • Subjects: Chemical engineering--Periodicals; Al(OH)3--ANFIS--Bayer process--Decomposition; Dewey: 660.205
  • Rights: legaldeposit
  • Publication Details: Taylor & Francis, Talylor & Francis Group
  • Abstract: Abstract :

    This article presents the results of nonlinear statistical modeling of the decomposition process of sodium aluminate solution, as part of the Bayer technology for the production of alumina. Based on the data collected in 2011 and 2012 from industrial production in the Birač Alumina Factory, Zvornik (Bosnia and Herzegovina), nonlinear statistical modeling of the industrial processes was derived. The model was developed as an attempt to define the dependence of the degree of decomposition of sodium aluminate solution as a function of the input parameters of the leaching process: caustic ratio ( α k) of the solution; ratio of the crystallization; content of Na2O(caustic)in the solution; the initial temperature of the solution; the final temperature of the solution; average diameter of the crystallized seeds; and duration of the crystallization process. As a tool for statistical modeling, Adaptive Network Based Fuzzy Inference System (ANFIS) was applied. The defined model using ANFIS methodology expressed a high level of fitting, and could be used to effectively predict the degree of decomposition of the sodium aluminate solution as a function of the input process under industrial conditions.


  • Identifier: System Number: LDEAvdc_100033795906.0x000001; Journal ISSN: 0098-6445; 10.1080/00986445.2015.1136292
  • Publication Date: 2016
  • Physical Description: Electronic
  • Shelfmark(s): ELD Digital store

Searching Remote Databases, Please Wait