Control Charts with Runs Rules for Poisson Process Data

摘要

We use Markov chains to compare run lengths of Poisson process individuals control charts with and without runs rules. Evidence quantifies the advantage of runs rules for certain cost structures z = C_α/C_β, where C_α is the cost of a Type I error, and C_β is the cost of a Type II error, and different shifts from in-control parameter λ_1 to out-of-control parameter λ_2.