Inverse Gaussian Control Charts for Monitoring Process Variability

Abstract

Control charts for detecting shifts in process variability are constructed under the assumption that the quality characteristic under study follows the inverse Gaussian IG(μ, λ) distribution with known parameter λ, and the location parameter μ is either known or unknown. The effects of the estimated probability limits on the performance of the proposed charts in detecting shifts in process variability are examined for the case when the parameter λ is unknown and must be estimated from preliminary subgroups taken from an in-control IG( μ, λ) process. The correction factors for evaluating the adjusted probability limits of the proposed charts that have a desired false alarm rate are given as well.

Keywords:

• Statistical process control
• Adjusted probability limits
• Run length distribution
• Average run length