Simple fuzzy models for prediction of flotation indices
Loading...
Date
2023
Authors
Jovanović, Ivana
Conić, Vesna
Sokolović, Jovica
Kržanović, Daniel
Radulović, Dragan
Journal Title
Journal ISSN
Volume Title
Publisher
University of Belgrade, Technical Faculty in Bor
Source
Proceedings - XV International Mineral Processing and Recycling Conference, IMPRC, 17-19 May 2023, Belgrade, Serbia, 2023, 404-409
Volume
Issue
Abstract
This paper presents the development and validation of two simple copper flotation models based on fuzzy logic (Mamdani and Takagi-Sugeno fuzzy inference system). Given that the Cu flotation process contains a large number of variables (especially inputs), models are called simple, because they contain only three input and three output variables. Input variables are feed grade, collector consumption in the roughing stage and overall frother consumption. Output variables are final concentrate grade and recovery as well as final tailings grade. The training and evaluation of the proposed models were accomplished on the basis of real process data from the industrial flotation plant of “Veliki Krivelj Mine”. The results showed that the proposed fuzzy models well describe the behavior of the industrial flotation plant in a wide range of circumstances (correlation coefficient R > 0.89 in all cases). There is almost no difference between the results, whether Mamdani and Takagi Sugeno fuzzy inference system is applied.
Description
Keywords
Fuzzy model, Flotation, Copper ore
Citation
DOI
Scopus
ISSN
ISBN
978-86-6305-133-1
License
CC-BY-NC-ND