小子样复杂系统保守可靠性增长模型

航空航天产品多为小子样复杂系统，其寿命分布服从双参数Weibull分布。由于样本量较小，Weibull分布的形状参数难以通过少数几次失效数据由客观方法估计得到。现有的处理方法是由专家设定或通过非参数统计方法解决，这就造成了可靠性分析结果具有很大程度的主观不确定性。为了研究这类系统的可靠性增长规律，文章提出了一种在不对其形状参数作任何假设的情况下的可靠度的保守估计方法，并进一步建立了系统分阶段研制过程中的可靠性增长保守模型，最后通过一个实例说明了该方法在工程中应用的有效性。

The Conservative Reliability Growth Model for Complex System with Small Samples

The Weibull distribution with two parameters is widely used in Aeronautical and Astronauticd products. For complex system with small samples, the shape parameter of Weibull is neither given by a few disable data nor estimated by impersonal method. At present, slove the problem by experts enacting or non-parameter statistic, which is subject to considerable uncertainty. This paper proposed an object and conservative method to estimate the shape parameter, at the same time get the reliability growth model for the system which is developed grading. At last, an example is given to illustrate the usage of the proposed method in engineering application.