APPLICATION OF HOTELLING'S T2 MULTIVARIATE STATISTICS CHART IN PRODUCTION CONTROL

Abstract

A multivariate statistical process control chart is a popular tool applied in quality management to control the production process. This paper introduced a Hotelling T² multivariate statistical chart that is applied to evaluate the lead-zinc flotation process according to the announced data. The method of building the Hotelling T² chart is shown through two stages. The results of control limit values and outliers were determined by the Hotelling T2 chart and the principal component analysis (PCA), performed by Minitab 21 software. Besides, the main parameters that affect the process were also analyzed by PCA and T2 decomposition methods. Like the univariate graph X, the Hotelling T² chart also identified “pulp level” and “upstream pH value” as the two crucial factors affecting the flotation process. However, the number of outliers on the Hotelling T² graph is less than those of the X chart that displayed the correlation of the parameters affecting the process efficiency. It can be said that the Hotelling T² multivariate chart will be a useful tool for statistical control of the production process owning many correlated parameters.