Application of Weibull distribution for assessments of space car reliability

Abstract

In the study of reliability, depending on the type of products, parametric and nonparametric models are used. Parametric models work with censored samples, where data for calculating reliability are presented as the operating time of failed and working products. The operating time of failed products is the time during which they failed during operation, and the operating time of working products is the time that they successfully worked. Usually, three basic laws of the distribution of earnings to failure are considered — exponential, Weibull and normal. Weibull’s law of distribution is most often used, since exponential is a special case of it and allows you to control the dynamics of reliability changes. Nonparametric analysis has the advantage of obtaining reliable results without reference to a specific distribution of lifetime data. However, due to this flexibility, nonparametric methods can be difficult and less convenient for solving various tasks typical of engineering design. At the same time, parametric analysis can more clearly identify specific trends and patterns of failures, which makes it preferable in some situations. Let’s consider and analyze the application of a nonparametric model for evaluating the reliability of spacecraft using the example of a generalized Way-bull distribution.