A fuzzy expert system application to the evaluation of ceramic- and paper-quality kaolin

This paper describes the results obtained from the application of a fuzzy expert system to the zonification of a deposit in terms of kaolin of a quality suitable for the paper and ceramic industries. In the first stage of the study, mineralogical, chemical and paper quality indices were characterized for samples taken from a kaolin deposit located in Galicia in NW Spain. In the second part of the study the expert system was trained and adjusted on the basis of analytical data, and its reliability in classifying samples in different quality groups was tested.Data analysis and the design and training of a fuzzy-logic expert system furnish the tool not only with general information inherent to kaolin mining but also with an input in the form of the local characteristics of the studied deposit. This method enables mining experts to identify the areas of a deposit containing kaolin that meets the criteria necessary for enrichment and subsequent use in the ceramic or paper industry. This kind of tool is not only indispensable in planning extraction tasks, but also permits technical and financial risk to be minimized. The model is innovative in that it automatically classified kaolin quality in terms of saleability, thereby emulating the logical reasoning processes of the human expert.