Generalized linear model based control charts with poisson response

The control chart is an important tool in statistical process control. Detecting the false alarm quickly, the control chart has a good performance and it is important in industrial processes. Generalized Linear Model based control charts use response variables that follow the exponential distributions. Performance of control charts are obtained by using MATLAB R2015a. The deviance residuals obtained from Poisson regression model are used as a control chart statistics. It is determined through simulation results that the performance of CUSUM control chart is the best. Shewhart control chart shows a good performance when the mean is shifted. Different from other control charts, EWMA control chart has worst performance in both cases.The plastic plywood data used in real case study. The CUSUM control chart quickly detects out-of-control observations.