Article ID: | iaor1998617 |
Country: | Australia |
Volume: | 15 |
Issue: | 2 |
Start Page Number: | 2 |
End Page Number: | 11 |
Publication Date: | Jun 1996 |
Journal: | ASOR Bulletin |
Authors: | Nicholls Miles G., Rodrigo Renuka |
Keywords: | fuzzy sets, mineral industries |
In this paper a modus operandi is discussed for the development of appropriate models of an aluminium reduction cell (RC), which incorporate all the significant aspects of the electrochemical relationships as well as the existing body of qualitative knowledge. The type of model that will be used in an RC will depend on its nature with respect to reliability of its operation, data reliability and availability. The purpose of the eventual model is to facilitate a better understanding of the operations of a reduction cell in a real time industrial situation, allowing for the determination of the optimal environment for the maximisation of the efficiency of the RC. Mathematical models have been developed in the past to represent micro aspects of the RC and recently a model was produced that approximated the operation of a reduction cell in a laboratory situation. However, in the ‘real world’ much work is yet to be done before such a reality materialises. In the approach to modelling the RC adopted in this paper, both the problem of highly questionable data and capturing the ‘soft’ knowledge of RC operations are tackled by the use of an object oriented methodology before model estimation and validation commences.