Statistical inference for assessing lifetime performance index of products with the Rayleigh distribution based on type II multiply censored sample

• Main Authors: Hung-Cheng, Hsu, 徐宏成
• Other Authors: Jong-Wuu, Wu
• Format: Others
• Language: en_US
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/83089376653690478047

碩士 === 國立嘉義大學 === 應用數學系研究所 === 99 === Technological development has been rapid in recent years. With the advent of modern technology, the manufacturers note that quality control (or management) becomes very important today. Thus, process capability indices (PCIs) have been widely utilized to assess whether product quality meets the required level. Here, we discuss the lifetime performance index (or larger-the-better process capability index) that is frequently used to measure the performance of process or product potential in manufacturing industries, where is the lower specification limit. Most process capability indices assume the product quality characteristic comes form normal distribution. But it’s not exactly true and sensible assumption in practice. Because of many lifetime of products generally may yield to non-normal distribution, such as exponential, Rayleigh, Gamma and Weibull distribution, etc. In addition, the lifetime data of product was missed for some reasons such as time limitation, budget constraints or material resources limitation, experimental difficulties, or artificial error of typist or recorder. Hence, taking the censored sample is more in keeping with real situation in the experiments. Thus, this study constructs approximate maximum likelihood estimator (AMLE) of the lifetime performance index based on the type II multiply censored sample from Rayleigh distribution. Then the AMLE of lifetime performance index is used to develop the new hypothesis testing algorithmic procedure based on type II multiply censored sample from the Rayleigh distribution in the condition of know . Finally, we give two examples to illustrated use of employ the testing procedure.