استعمال أسلوب بيز للوحات سيطرة لمجموع المتراكم

الملخص EN

In this study, we will introduce a Bayesian approach for The usual operation of the cumulative sum(Cusum) chart.

The Cusum control chart used for monitoring a process mean when the observations (product) follows Binomial Distribution in continuous manufacturing.

By assuming that the prior information available about the parameter p (the proportion of defective items in the production) are a beta distribution, with different values for the parameters (a, b) and the uniform distribution over the [0, 1] interval.

we derived the posterior distribution of p and the average run length (ARL) and standard deviation run length (SDRL) will be determined by using a Bayesian method.

The results were obtained by using Programs written using matlab-R2018a program.

The results shown that Cusum control charts, when the prior information on p is a beta distribution (a, b) were more sensitive at certain values for the parameters a, b according to the smallest values of ARL and SDRL.

In general, we can see that Cusum control charts, when the prior information on p are a beta distribution (a, b) and the uniform distribution over the [0, 1] interval were more sensitive, when we adopt new starting value and by increasing the number of states for markov chain and reducing the decision interval.