DIAGNOSING FOR SIGNAL IN MULTIPLE CORRELATED PROCESSES

Recently,considerable attention has been given to the effect of data c orrelation on statistical process control (SPC). Use of traditional SP C methods when observations are correlated often leads to misleading c onclusions for diagnosis of processes. A good solution to the problem of diagnosis for multiple correlated processes should be that it not o nly can be used in monitoring the processes,but also is able to diagno se the behavior of each process. In terms of the above meaning require d,this problem has been opened for many years. In this paper,an attemp t to solve the problem is made. A procedure is systemly reported by re gression residual for multiple correlated processes when an singnal of out -of-control occurs and an example of implementatim of the procedu re is given.