Development of cumulative sum control controls and dual rhythm moving moving averages based on generalized linear regression to monitor cascading processes
Volume 7, Issue 2, Summer 2017, Pages 82-93
Shervin asadzadeh, Ali Torabi
Abstract Nowadays, the process of producing most products is such that the products are created in different and interdependent stages. . One of the most widely used and specialized control charts used to monitor multi-stage processes is the selection diagrams of deviant agents. These control charts have been further developed and used for high quality characteristics. In this paper, two control diagrams based on generalized linear patterns are proposed to monitor a two-step process with a two-sentence qualitative characteristic in the second step. In the proposed approach for monitoring binomial variables, a cumulative sum control control diagram and a dual-balance moving average control diagram are designed considering the relationship between qualitative characteristics in two stages. The proposed control diagrams are based on the residual values of the deviation. In order to establish the relationship between the qualitative characteristics of the first and second stages, a new interface function has been used in the framework of the generalized linear pattern. The performance of the proposed control diagrams is evaluated using simulation based on the average sequence length criterion. The results show that the ability to detect cumulative summand control charts and the proposed dual rhythmic moving average is far better than the control charts in the literature.