基于组合设计的软件可靠性测试方法

提出了将实验设计中的组合设计方法应用于软件可靠性测试。分析了组合设计方法在软件可靠性测试用例设计中应用的理论基础和基于组合设计的软件可靠性测试的一般方法。同时指出了利用软件操作剖面信息和失效数据进行可靠性评估的方法。     

应力-强度模型的Bayes可靠性分析

当应力、强度分别服从于正态分布、指数分布和Weibull分布时 ,分析了应力 -强度模型的可靠性评估 ,着重讨论了无信息验前下的Bayes可靠性评估。仿真结果表明 ,无信息验前下的评估结论可以很好地用频率学派的观点来解释。     

基于神经网络的弹丸强度可靠性分析

构造了一个基于神经网络的弹丸强度分析模型,通过对以往实际发生数据的训练与推理,设计并构造了一种自适应变步长自构型神经网络。对影响弹丸强度的诸多因素进行了综合的非线性分析。并可以利用该模型进行强度预测,减少实际打弹的次数,经济效益明显。     

火箭发动机可靠性增长评估的Bayes方法

针对火箭发动机分阶段试验,每阶段试验后对发动机出现的故障分析、归零的情况,提出等效试验数据的概念,结合Bayes方法对系统的可靠性增长试验进行评估,并在此基础上给出可靠度的增长分析,最后给出实例进行说明。     

Inventory competition with yield reliability improvement

This article studies the inventory competition under yield uncertainty. Two firms with random yield compete for substitutable demand: If one firm suffers a stockout, which can be caused by yield failure, its unsatisfied customers may switch to its competitor. We first study the case in which two competing firms decide order quantities based on the exogenous reliability levels. The results from the traditional inventory competition are generalized to the case with yield uncertainty and we find that quantity and reliability can be complementary instruments in the competition. Furthermore, we allow the firms to endogenously improve their yield reliability before competing in quantity. We show that the reliability game is submodular under some assumptions. The results indicate that the competition in quantity can discourage the reliability improvement. With an extensive numerical study, we also demonstrate the robustness of our analytical results in more general settings. © 2015 Wiley Periodicals, Inc. Naval Research Logistics 62: 107–126, 2015     

利用验前信息的一种序贯检验方法——序贯验后加权检验方法

序贯验后加权检验方法是Bayes 统计观点在假设检验中的运用。本文提出了该方法的一般理论,给出了当总体的分布参数具有验前信息时的序贯检验方法。文中确定了决策区的划分,同时讨论了序贯截尾方案,给出了检验中可能犯两类错误的概率的上界,并将一般理论应用于产品的可靠性检验和再入飞行器随机落点的精度鉴定。由于运用了验前信息,因此能有效地在少量试验之下进行统计假设检验。本文所提供的方法对于昂贵产品试验结果的统计评定具有普遍意义。     

Testing or fault-finding for reliability growth: A missile destructive-test example

A new piece of equipment has been purchased in a lot of size m. Some of the items can be used in destructive testing before the item is put into use. Testing uncovers faults which can be removed from the remaining pieces of equipment in the lot. If t < m pieces of equipment are tested, then those that remain, m₁ = m − t, have reduced fault incidence and are more reliable than initially, but m₁ may be too small to be useful, or than is desirable. In this paper models are studied to address this question: given the lot size m, how to optimize by choice of t the effectiveness of the pieces of equipment remaining after the test. The models used are simplistic and illustrative; they can be straightforwardly improved. © 1997 John Wiley & Sons, Inc. Naval Research Logistics 44: 623–637, 1997     

MANET多路径路由中最大可靠性路径选择算法

如何选择路径的数量和质量对多路径路由机制的性能有着重要的影响。已有的多路径算法没有深入研究如何选择多路径的问题。对目前存在的两个典型问题进行了分析,在此基础上研究了路径可靠性模型和虚拟完全非交叉多路径模型,然后提出一个最大可靠性多路径选择算法。算法利用路径权重作为路径可靠性的近似解决方案,以此克服路径可靠性度量问题(NP难题)研究的复杂性,根据路径可靠性模型和完全非交叉多路径模型来选择可靠的路径集,使用这组路径集并行分布流量。应用OPNET模拟平台实现了算法,结果表明,本算法能增加聚合带宽,优化网络带宽的应用,提高网络的吞吐率和多路径路由的性能。     

On a class of multiple failure mode systems

The primary objective of this work is to introduce and perform a detailed study of a class of multistate reliability structures in which no ordering in the levels of components' performances is necessary. In particular, the present paper develops the basic theory (exact reliability formulae, reliability bounds, asymptotic results) that will make it feasible to investigate systems whose components are allowed to experience m ≥ 2 kinds of failure (failure modes), and their breakdown is described by different families of cut sets in each mode. For illustration purposes, two classical (binary) systems are extended to analogous multiple failure mode structures, and their reliability performance (bounds and asymptotic behavior) is investigated by numerical experimentation. © 2002 Wiley Periodicals, Inc. Naval Research Logistics 49: 167–185, 2002; DOI 10.1002/nav.10007