Automated control of classifocation in a hydrocyclone with incomplete information

Abstract

The main areas of achievement of economic benefit in iron ore enriching is to increase the productivity of production units and improving the quality of the product, which requires a complex automation of enrichment processes. Control of processing complex is costly – expensive measuring equipment and considerable computing power. In addition, iron ore beneficiation process should be viewed as a distributed system consisting of separate processes with separate control systems, interconnected and influenced each other. Considering hydrocyclone of one single, the second stage of grinding, can greatly simplify the calculations and allow to consider possible reaction. It is advisable to use modern intelligent automated management tools – optimum and adaptive control, means of artificial intelligence, fuzzy logic, genetic algorithms, hybrid models. Studies show that the fuzzy control of hydrocyclone of second stage of grinding allows to take into account a lot of dependencies and develop the controlling influences, dependent on many parameters. In addition, this approach allows to work in the face of uncertain parameters. The presented control system learns and configures itself, as well as takes into account the link with the previous and subsequent grinding stage, affecting the overall distributed system. Further studies suggest a deeper study of the relationship between the mechanisms of the various grinding stages.