Case-based reasoning for crystallizer selection using rough sets and fuzzy sets analysis

Crystallization conditions in the crystallization unit and composition of the solution determine the crystal size distribution and shape. The objective of this study was to chart those physical properties and operating conditions that could explain the crystal size distribution for a compound in the industrial crystallizer using artificial intelligence based on the already existing data obtained for various compounds. The authors applied case-based reasoning (CBR) which is based on the reuse of the past experience in finding the solutions to new, similar problems. The proposed approach may save the time-consuming experimental work when the crystal size distribution of a new compound should be predicted. The example of continuous suspension crystallization is used to prove the usefulness of the proposed approach using rough and fuzzy sets in adaptation phase of case-based reasoning.