MONITORING THE TEMPERATURE OF THE MOLD FOR THE MANUFACTURE OF STEEL INGOTS USING REGRESSION RESIDUALS THROUGH CONTROL CHARTS
The study aims, monitor 230 observations in the temperature of the mold for the manufacture of
steel billets through the waste from the regression analysis by means of the control chart X-bar and
R. Whit the values of mold temperature, initially estimated the equations independently by the
method of ordinary least squares, analyzing the individual variable. With waste from the equations,
we calculated the correlation matrix, which showed that significant correlation exists at the level of
significance. As the data analysis are presented interrelated, it is not possible to directly use these
values in the control charts then determined a mathematical model that enabled a joint estimation,
considering the relationship between the variables and relationships that occur between them,
obtaining an equation for each variable to then apply the control charts in the waste that model.
The mathematical model waste fitted using the simple dynamic regression model, under which the
parameters were obtained statistically significant, and waste analysis showed that the fitted model
was appropriate. In the verification step of the process stability was found that all points are within
the control limits.
PaperID: p 030-044
Author's Name: Vanusa Andrea Casarin, Regiane Klidzio, Adriano Mendonça Souza and Jaime Alvares Spim
Volume: Volume 3
Issues: Volume 3
Keywords: Regression analysis; control charts; residues.