Capability Sampling Plans For Acceptance Or Rejection Of A Production Lot With The Lifetime Data Binding An Exponential Distribution
The quantifications of parameter location and variation are a central for determining variable types of acceptance sampling plans to understanding the quality of the process. Practically, the estimators of unknown parameters are not unit-less, which are inconvenient summary statistics in a plant or supply base where a variety of characteristics with disparate metric measures are considered. The dimensionless quantities of process capability indices, measuring actual performance of an in-control process with regard to the specification limits, have played an integral role in continuously improving the quality and reliability of products. However, most indices’ estimators having been applied so far assume measurements of the quality characteristic are normally distributed. The exponential distribution is commonly utilized to model the electronics components and systems, mechanical fatigue failures, and some corrosion processes that usually do not wear out until long after the expected life of the product being installed. In this paper, based on the lifetime capability index, we optimize acceptance sampling schemes for the exponential population with/without censoring by employing statistical and decision-theoretic methodologies that minimize the number of failures needed on the inspection. Keywords: Sampling plans, Life capability, Producer risk, Consumer risk, Non-linear optimization.