多源信息的Bayes融合精度鉴定方法

为了合理运用多种验前信息 ,文中首先引入了验前信息的可信度的概念 ,由此论述了验前信息与现场试验信息的融合方法 ,给出了多源信息下的Bayes精度鉴定方案。最后以实例说明鉴定方案的运用。     

Variables inspection for SPC‐quarantined production

When a control chart signals an out‐of‐control condition for a production process, it may be desirable to “quarantine” all units produced since the last in‐control SPC sample. This paper presents an efficient procedure for variables inspection of such “SPC quarantined” product. A Bayesian sequential inspection procedure is developed which determines whether the out of control production is of acceptable quality. By inspecting the units in reverse of the order in which they were produced, the procedure is also capable of detecting the point at which the process went out of control, thus eliminating the need to inspect units produced prior to the onset of the out of control condition. Numerical examples are presented, and the performance characteristics of the procedure are demonstrated using Monte Carlo simulation. © 2001 John Wiley & Sons, Inc. Naval Research Logistics 48: 159–171, 2001     

Bayesian testing strategies for software with an operational profile

We consider a software reliability model where the failure rate of each fault depends on the specific operation performed. The software is tested in a given sequence of test cases for fixed durations of time to collect data on failure times. We present a Bayesian analysis of software failure data by treating the initial number of faults as a random variable. Our analysis relies on the Markov Chain Monte Carlo methods and is used for developing optimal testing strategies in an adaptive manner. Two different models involving individual and common faults are analyzed. We illustrate an implementation of our approach by using some simulated failure data. © 2001 John Wiley & Sons, Inc. Naval Research Logistics 48:747–763, 2001     

初始先验分布未知条件下的DLMR及其Bayes预测

讨论了具有无信息初始先验分布的矩阵变量动态线性模型 (简记为DLMR)及其Bayes预测 ,利用MDIP方法 ,就尺度方差阵∑已知和未知两种情况 ,分别给出模拟初始状态参数(θ1 |D0 )和 (θ1 ,∑ |D0 )的无信息先验分布 ,进而给出其Bayes预测。     

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     

Stochastic production capacity: A bane or a boon for quick response supply chains?

The quick response (QR) system that can cope with demand volatility by shortening lead time has been well studied in the literature. Much of the existing literature assumes implicitly or explicitly that the manufacturers under QR can always meet the demand because the production capacity is always sufficient. However, when the order comes with a short lead time under QR, availability of the manufacturer's production capacity is not guaranteed. This motivates us to explore QR in supply chains with stochastic production capacity. Specifically, we study QR in a two-echelon supply chain with Bayesian demand information updating. We consider the situation where the manufacturer's production capacity under QR is uncertain. We first explore how stochastic production capacity affects supply chain decisions and QR implementation. We then incorporate the manufacturer's ability to expand capacity into the model. We explore how the manufacturer determines the optimal capacity expansion decision, and the value of such an ability to the supply chain and its agents. Finally, we extend the model to the two-stage two-ordering case and derive the optimal ordering policy by dynamic programming. We compare the single-ordering and two-ordering cases to generate additional managerial insights about how ordering flexibility affects QR when production capacity is stochastic. We also explore the transparent supply chain and find that our main results still hold.     

战斗部虚拟试验贝叶斯线性校准预测方法

聚能装药战斗部射虚拟试验中,射流穿深对靶板材料模型参数较敏感。为建立较准确实用的射流穿深预测模型,提出一种贝叶斯线性校准方法,建立物理过程真实值、模型预测值和试验测量值间的统计关系,结合有限的试验数据,校准未知模型参数。实例分析表明,该方法可较好地解决包含各种不确定性的大型复杂计算模型的参数校准问题。     

Optimal stopping in software testing

In this paper we address the problem of how to decide when to terminate the testing/modification process and to release the software during the development phase. We present a Bayesian decision theoretic approach by formulating the optimal release problem as a sequential decision problem. By using a non‐Gaussian Kalman filter type model, proposed by Chen and Singpurwalla (1994), to track software reliability, we are able to obtain tractable expressions for inference and determine a one‐stage look ahead stopping rule under reasonable conditions and a class of loss functions. © 2002 Wiley Periodicals, Inc. Naval Research Logistics, 2003     

多阶段含延缓纠正的可靠性增长序化建模分析

多阶段含延缓纠正的可靠性增长试验,其各个阶段内的试验数据分别服从参数不同的Weibull过程.针对现场样本少、传统的可靠性评估方法实施困难的问题,首先在获得各阶段形状参数的基础上,对各个阶段截尾时刻的失效强度建立顺序约束关系.然后,运用Bayes方法融合先验信息,获得最末阶段截尾时刻的失效强度的期望与置信上限.该结果融合了历史阶段的样本信息,优于仅利用单个阶段样本的分析结果,适用于多阶段可靠性增长试验的综合评估.     

基于贝叶斯网络的装备损伤定位系统研究

分析了贝叶斯网络在装备损伤定位方面的优势,及其损伤定位的方法与流程,建立了用于装备损伤定位的贝叶斯网络模型,并开发了损伤定位系统。以某型火炮为例,演示了其损伤定位的一般过程,验证了贝叶斯网络在装备损伤定位中应用的可行性与有效性。