An aging-intensity-function-based parameter estimation method on heavily censored data

Building life distribution models of key components of a product is a basic reliability problem. When the data for modeling are heavily censored, the conventional estimation methods such as maximum likelihood method and least square method are no longer applicable. Several approaches have addressed this issue and innovative methods are still needed. This paper presents such a method. It is based on aging intensity function (AIF) of a distribution. The proposed method consists of two main steps. In the first step, a semi-parametric method is used to estimate the empirical AIF, which is further transformed into the shape parameter values. These values are then fitted to a distribution, from which the point and interval estimates of the shape parameter are obtained. In the second step, the point estimate of the shape parameter is fixed and a single parameter maximum likelihood method is used to estimate the scale parameter. Two examples that involve six datasets are included to illustrate the performance of the proposed method. The results show that the proposed method has an excellent performance in terms of simplicity, generality, accuracy, and unbiasedness.

Automated summary

This paper proposes a method based on the aging intensity function (AIF) to build life distribution models for key components of a product. The method utilizes a semi-parametric approach to estimate the empirical AIF and obtain point and interval estimates of the shape parameter, and then uses a single parameter maximum likelihood method to estimate the scale parameter. The results from two examples with six datasets demonstrate the excellent performance of the proposed method in terms of simplicity, generality, accuracy, and unbiasedness.