复杂可修系统可信度建模中任务维修度计算的统计推断方法

STATISTICAL INFERENCE TO MISSION MAINTAINABILITY IN DEPENDABILITY MODELING FOR GROUND VEHICLE SYSTEMS

在设定每一部件的致命性故障率为常数并在确定出各部件任务修复时间置信上限的基础上，通过对可能发生的多次致命性故障事件的概率状态描述，得到了部件故障率与系统可信度的定量关系，进而依据这一可信度的建模原理归纳出不考虑任务修复时间的具体分布类型时复杂可修串联系统可信度的计算方法，同时也给出了任务维修度的统计推断表达式；最后，以典型的 ４ 部件串联可修系统为例，探讨了致命性故障次数、允许任务修复时间、部件任务修复时间的不同置信上限等的变化对系统可信度的影响，并论证了系统可靠性、维修性和保障性与可信度的关系。

On the basis of assumed constant failure rate and determined repair time at upper α% point for each unit within complex repairable systems such as ground vehicles and automobiles, a mathematical approach to dependability modeling regardless of the probability density of maintenance time, has been developed by means of description of multiple critical failures to occur or having occurred in the operation of the complex repairable system, and the procedure of estimating mission maintainability with statistical inference is presented in the paper. Finally, the effect of the time allowable for maintenance, malfunction number and repair time at upper α% point for each unit upon system dependability is explored through the example of a four unit repairable series system, and we quantificationally analyze in sufficient detail the relationship of dependability between reliability, maintainability and supportability, which could be used to the system configuration optimum and recovery from malfunction. The approach proposed in the paper would give a new basis for dependability modeling.