含共因事件航天系统的通用动态故障树方法

为了表示航天复杂系统中组件依赖关系和输入次序等动态事件与共因事件，动态故障树的求解过程通常会面临很高的计算复杂度与时间开销。基于系统定义假设，由基本事件的概率分布入手，根据动态门的时序逻辑定义和关联方式，给出系统故障事件的准确通用计算方法GDFTA (Generic Dynamic Fault Tree Algorithm)。GDFTA方法实现了通用的动态树求解，避免了Markov方法的状态空间爆炸问题，并取得了较理想的计算准确性。通过对动态树实例的可靠性评估证明该方法可行。在标准动态故障树测试集中，四类不同系统的可靠性结果精度优于Monte Carlo方法，计算效率也有显著提高。

A Generic Algorithm for Dynamic Fault Tree in Spacecraft with Repeated Events

Dynamic fault tree approaches are generally used to address repeated dynamic events such as component dependence and order of input in space complex systems, but its solving process often lead to ultra high computational complexity. An accurate GDFTA (Generic Dynamic Fault Tree Algorithm) method based on system definition assumptions is used to analyze the type of failure. This method begins with the probability distribution of basic events and takes into account the logical time sequence and the connection style of the dynamic gates. The GDFTA outperform the Markov method in solving the universal dynamic fault trees for avoiding the Markov state space crisis. In experiments, several dynamic tree examples are used to test its feasibility. Then, experiments using the standard dynamic tree benchmark set show high accuracy of reliability test results of four types of systems. Moreover, the computational time is much less than a Mante Carlo simulated method.