A Study On Quality Control Charts Effect For Autocorrelated Process

Statistical process control is mainly on the use of various types of control charts to monitor the process to ensure that the process is in a steady state. When the process is in abnormal situation, control charts should be send the alert timely. Traditional control charts based on the premise that the process of building plans independent and identically distributed, but the presence of self-related phenomena often exist in the reality. When the process is autocorrelation, the use of traditional control charts for process monitoring will send a large number of false alarms issued, to produce a tremendous loss.Paper study the process which exist the autocorrelation phenomena. Firstly, the autocorrelation function and partial autocorrelation function to determine whether there is autocorrelation process, when the process is autocorrelation, using the corresponding time series model to fit the quality of property values, and then obtain the residuals. By normality test and autocorrelation function, partial autocorrelation function verification, there is no residual sequence derived from the relevant phenomena, in line with the traditional control charts drawing premise. Then the residual sequence established for the study of a variety of control charts.This paper studies the AR (1) time series model, and in Chapter4of the simulation process. In the course of time in a stable and non-stationary, this paper the corresponding AR (1) model in different autocorrelation coefficients do simulation, and control charts to make a single value, residual control chart, residuals and residual EWMA control chart Poor CUSUM control chart. Through comparative studies, the conclusion that the residuals and residual EWMA control chart CUSUM control chart for small shifts in the process mean particularly sensitive. In Chapter5of this paper, through the study of examples of the analysis, further proof of the reliability of conclusions.Because the residuals and residual EWMA control chart CUSUM control chart can make full use of historical data, for data collection and data less difficult production process is particularly applicable. Therefore, this paper suggested many varieties of small batch production process and the use of residual residual EWMA chart CUSUM control chart for process monitoring.