This work proposes a method for the application of regression control charts in the monitoring of manufacturing processes. The proposed method is presented in two phases: retrospective analysis (Phase I) and process monitoring (Phase II). It includes a simple modification of the multiple regression control chart, allowing the monitoring of the values of quality characteristics of the process, instead of monitoring the regression standardized residuals. It also proposes an extrapolation control chart, which verifies whether the control variables extrapolate the set of data used in regression model estimation. The proposed method was successfully applied in a rubber manufacturing process. The Average Run Length (ARL) distribution was estimated using the Monte Carlo method, proving the efficiency of the proposed chart in detecting some alterations in process parameters.
Regression control chart; Linear regression models; Average run length
