基于信息融合的系统可靠性优化试验设计模型

发布时间：2018-01-21 04:29

本文关键词： 可靠性试验设计信息融合 Bayes方法验后风险准则风险计算　出处：《航空动力学报》2016年02期 　论文类型：期刊论文

【摘要】：分析了现今可靠性试验设计的不足,考虑两类风险(弃真风险、采伪风险)建立了基于信息融合的可靠性优化试验设计模型(ORTIF,optimization reliability test design modeling based on information fusion).分析了可靠性试验设计的需求、约束条件,基于子系统和系统验前分布是Beta分布,提出了可靠性系统中子系统层与系统层之间的信息融合技术.根据验后风险准则,给出弃真和采伪风险的定义.基于Matlab软件给出了最优试验方案数值计算的求解步骤.最后以液态火箭发动机为例,求解出满足约束条件的最优试验方案为(9,5,10,6,1),对应的最小试验费用为3 326.9,并得出权值方案对试验方案的确定影响较大的结论.
[Abstract]:The deficiency of reliability test design is analyzed, and the ORTIF model of reliability optimization test design based on information fusion is established by considering two kinds of risks (forgery risk and forgery risk). Optimization reliability test design modeling based on information. Fusioning). The requirements of reliability test design are analyzed. Based on the fact that the prior distribution of subsystems and systems is Beta distribution, the information fusion technology between subsystem layer and system layer in reliability system is proposed. The definition of the risk of forgery and forgery is given. Based on the Matlab software, the steps of numerical calculation of the optimal test scheme are given. Finally, the liquid rocket engine is taken as an example. The optimal test scheme to satisfy the constraint conditions was obtained, and the minimum test cost was 3 326.9. It is concluded that the weight scheme has a great influence on the determination of the test scheme.
【作者单位】：国防科学技术大学信息系统与管理学院;
【基金】：国家自然科学基金(70971133)
【分类号】：E92;TB114.3
【正文快照】： 随着系统复杂化与集成化的发展,系统试验周期越来越长,尤其是对军事装备而言.复杂装备系统试验可提供的现场试验样本量越来越小,甚至出现极小子样的情况(如神舟飞船的交会对接试验).相对于系统层试验数据少的不足,子系统存在着大量的相关的试验数据,如子系统试验数据、仿真试

【参考文献】