This paper shows two applications of artificial intelligence models in process development. In the first application, a heuristic fuzzy correlation can predict flotation behavior in bench scale based on feed characteristics for a high variability copper ore, as well as to predict plant performance for a feasibility project evaluation, minimizing its risk. The second case study shows the use of neural networks to model the iron ore sintering, correlating ore characteristics and operational parameters with sinter properties. With this application, CVRD aims to establish a new step on iron ore product evaluation and customer support. |