Aplicação e análise dos métodos soma cumulativa e média móvel exponencialmente ponderada em dados autocorrelacionados para o monitoramento e controle do processo de flotação de minério de ferro Tanckcell

Abstract

In iron ore processing plants, flotation is an essential process to achieve desired quality standards, in which efficient management of controllable variables can be obtained by monitoring small variations in the silica content of the concentrated mineral. The objective of this study was the analysis and comparison of control charts of variables, using graphs of cumulative sums and exponentially weighted moving average, to determine the most viable application in monitoring the process, with the nature of autocorrelated data. The methodological procedures used for this research were organized into stages, the steps for constructing the cards were defined, after evaluating the nature of the data with non-normal distribution, with tests and procedures, in order to interrupt the autocorrelation, proving the correct use of control charts. The results of this research showed that it was possible, by analyzing and comparing the two charts, to establish the indicators of each one and with greater capacity, to use the exponentially weighted moving average chart, to reduce the level of contamination in the concentrate, which directly influences in the quality of iron ore, a product of flotation.