Phase II monitoring and diagnosis of autocorrelated simple linear profiles

•Effective monitoring schemes are developed for autocorrelated profiles.•The proposed diagnostic aid helps users determine the change point of the process.•The proposed methodology can be applied in industry and other areas.

If the quality of a process is better represented by a functional relationship between response variables and explanatory variables, a collection of this type of quality data is called a profile. In this paper, we consider the functional relationship which can be represented by a simple linear regression model with a first-order autocorrelation between error terms. We propose exponentially weighted moving average (EWMA) charting schemes to monitor this type of profile. The simulation study shows that our proposed methods outperform the existing schemes based on the average run length (ARL) criterion. We also propose a maximum generalized likelihood ratio method to obtain a change-point estimator to help users determine the assignable causes.