Statistical Process Control For The Two-Component Model

The two-component model which incorporates both additive and multiplicative disturbances into a regression model has been found to accurately describe many industrial processes. In this paper,we discuss the SPC problem of the data that satisfied the R-L model.We obtain the estimation of the model parameters firstly. The maximum likelihood estimates (MLEs) of the model are computed with the EM algorithm. Then we use adaptive exponentially weighted moving average (AEWMA) control charts to control the process which satisfies the two-component model. When the input state level unchanges, the output of the process are independent each other. We design an AEWMA chart which can quickly detect small and large shifts in the output simultaneously. During the input state level changing process,the output are autocorrelated and have no tendency to respond immediately. Therefore we use another AEWMA method to predict the output, then we can monitor and controll the forecast errors as the solution of the control matter in the transition process.