步兵榴弹环境适应性试验设计与评估方法

直接由验前数据获取验前分布可能导致正样机数据"左右"定型试验结论,为了解决此问题,采用混合贝塔分布描述二项分布的验前分布,用相似性因子描述前后两次试验在产品状态和试验条件的差异,运用层次分析法确定了影响步兵榴弹环境适应性试验项目的不同因素的取值范围和相应评语集,给出了基于混合贝塔分布的试验方案设计方法.在获得试验数据后,运用聚类分析方法分析了产品验后分布的环境敏感性.结果表明,该方法参数估计的均方误差更小,估计值更稳健;聚类分析方法能够发现环境因素对产品性能的影响,能够为产品的改进和使用提供数据支撑和指导.     

基于贝叶斯原理的小子样试验导航误差评定方法

针对鉴定试验花费大、周期长的问题,提出一种水下无人航行器导航误差小子样贝叶斯试验CEP评定方法。该方法综合运用大量验前子样,对其进行相容性检验,以确定有效的验前子样,并运用统计推断理论获取验前信息。根据在验收试验中获取的验后子样获取验后信息,利用贝叶斯理论推出无人航行器满足CEP指标的概率。实例结果表明了方法的有效性和实用性。     

测量导致设备性能退化条件下的剩余寿命建模分析

测量本身导致设备性能退化的情况是存在的,然而这个问题往往被研究者们所忽视。文中针对这种情况下的剩余寿命建模进行了研究。首先,在自然退化模型的基础上建立有损测量退化模型;其次给出了相应模型的参数估计方法和基于贝叶斯理论的设备剩余寿命的实时更新方法。最后通过一个仿真示例证明了研究这个问题的意义和文中建模的有效性。     

基于贝叶斯估计的起重机械主梁的可靠性预测

针对起重机械主梁工作环境的复杂性和在相同工作环境下可靠性数据不足的问题,将贝叶斯估计方法和数据拟合相结合,以主梁的退化量作为性能参数,建立了起重机械主梁在工作环境下的可靠性预测模型;利用贝叶斯估计方法最大限度地将工作现场数据和历史试验数据相结合,及时更新主梁在工作环境下性能参数的分布函数,进而估计工作环境下不同时刻主梁的可靠度,最后采用最小二乘法估计主梁的可靠性函数.预测结果表明,起重机械主梁失效满足参数为(9.290,225.9)的2参数威布尔分布,得到为了满足主梁98％的可靠度要求,最小的维修间隔周期为149天.     

朴素贝叶斯分类模型在指导学生选择专业选修课方向中的应用

目前我国的课程体系体现"大专业、宽专业基础、多方向"的原则,但面对众多课程,学生对课程间的关联关系不甚清晰,对专业选修课方向的选择非常迷茫。本文充分利用以往各届学生选择专业选修课方向的先验知识,建立适用于专业方向选择的分类模型,对需要指导的学生根据他们的学习情况进行分类预测,帮助他们合理地选择专业选修课方向。     

战术导弹飞行试验的检验方法比较

检验方法对于战术导弹飞行试验与鉴定至关重要,介绍了定型与批检试验的检验方法及其发展变化,并对它们的特点及使用做了比较.     

An Approach to Combining Results From Multiple Methods Motivated by the ISO GUM

The problem of determining a consensus value and its uncertainty from the results of multiple methods or laboratories is discussed. Desirable criteria of a solution are presented. A solution motivated by the ISO Guide to the Expression of Uncertainty in Measurement (ISO GUM) is introduced and applied in a detailed worked example. A Bayesian hierarchical model motivated by the proposed solution is presented and compared to the solution.     

An approach for the Bayesian estimation in the case of ordered piecewise constant failure rates

In this paper we present an approach for the Bayesian estimation of piecewise constant failure rates under the constraint that the constant value of the failure rate in an interval of time is greater than a function of its values in the prior intervals. We apply this approach to the estimation of piecewise constant failure rates for conditional IFR, IFRA and NBU distributions. The prior distribution for the failure rate in each interval is specified through gamma distributions with functions of the failure rate values corresponding to the rest of the intervals as location parameters. Using this approach the prior distribution parameters have interpretations through prior means and variances of the values of the piecewise constant failure rate. The posterior distributions and expected values can be found in terms of gamma functions, without the necessity of numerical integrations. We apply this approach to a model for reliability estimation when two operational modes exists and the number of failures in each operational mode is unknown. Finally a numerical example is presented in which simulations of posterior densities are carried out.     

Notes on a recursive procedure for point estimation

Summary A procedure for making inference about parameters of a statistical model is proposed which uses a general and easy to implement recursive system. Unlike the a full Bayesian approach, this procedure does not require a particular choice of prior, only a suitable (mean, variance)-parametrised class. Unlike maximum likelihood, it takes into account the global structure of the likelihood, and seems reasonable even when pathological cases of the likelihood arise. An heuristic large-sample argument is also considered.