A Sample Truncation Rule for Product Reliability Estimation in Multiple Censoring Data Conditions

Abstract

This paper studies the method to estimate the location parameter of a three-parameter Weibull life distribution in the situation of multiple censoring life data available, the effect of sample size and minimal outcome on life and reliability estimation, and the policy to choose favorable data from operation records. First of all, a method to estimate the location parameter of the Weibull distribution is presented in the scenario of product probabilistic life and reliability estimation. Relationships between the estimated life distribution parameter and the right-censored life data applied is demonstrated, a rule is proposed to choose the appropriate number of censoring life data by discarding some shorter operation time records. It is illustrated that an appropriate combination of sample size and the minimum observation value (right-censored life time) is reasonable for the balance between estimation confidence level and reliability value.