刀具磨损监测的随机向量相关度量法

本文提出了一种运用功率累积和与n 维随机向量相关度量监测刀具磨损的新方法,即通过计算刀具三向振动加速度的相关度量系数,主轴扭矩或某项振动加速度的功率累积和综合监测刀具的磨损、破损以及刀具磨损的程度。该法在简单性、可靠性等方面比以往提出的方法有较大的进步,因而使得本方法具有很大的应用前景。文中对该法的理论基础进行了严格的证明,并给出了实验论证结果。     

原始变量守恒形式控制方程的时间准确性分析

基于压力、速度和温度的原始变量为自变量的守恒形式的控制方程可应用于定常流动问题，但是在求解非定常问题，例如某一典型激波管问题时，激波后温度出现过冲现象，即使通过细化网格、提高空间格式精度或者换用其他通量格式仍不能消除，这表明误差可能来自该方法本身。采用一维Euler方程对该方法进行数值分析。分析结果表明，数值误差来自时间项。通过构造相应的双时间步方程，虚拟时间项采用原始变量，而物理时间项采用守恒变量，并在两个相邻物理时间步内作为定常问题求解，可以收敛到相应的守恒形式，消除上述误差，得到准确的非定常数值解。     

几类特殊多叶函数的极值点和支撑点

设S(p,α)、C(p,α)及K(p,α)分别表示α阶的p 叶星像、凸像和近于凸函数族。本文对这些函数族的支撑点和S(p,α)的从属函数族的极值点和支撑点进行了研究,推广了Grassmann 和Perera 等人的结果。     

Bernoulli抵消策略的负顾客的M/G/1排队系统

负顾客的排队模型是排队论近年来的新兴分支。文中引入了带有B ernou lli抵消策略的负顾客的M/G/1排队模型,即到达的负顾客以概率p抵消队首正在接受服务的正顾客,以1-p的概率抵消队尾的正顾客,利用补充变量法,状态转移法和L变换,最后得到系统稳态队长的概率母函数。     

基于混合变量传递函数方法的圆柱壳屈曲分析

提出了一种分析正交各向异性圆柱壳和阶梯圆柱壳稳定性问题的混合变量传递函数方法。首先在壳体环向利用三角级数对位移变量进行展开 ,利用Fl櫣ｇｇｅ薄壳理论和变分原理 ,建立圆柱壳的平衡方程 ,找出对偶力变量 ,将平衡方程写成混合变量形式 ;通过定义混合状态变量 ,建立了系统的状态空间方程 ;然后利用传递函数方法 ,得到了具有任何轴对称边界条件轴压圆柱壳屈曲问题的解析解 ;最后通过位移连续和力平衡条件 ,得到了阶梯圆柱壳屈曲问题的解。理论解推导过程表明此方法在引入边界条件和进行阶梯圆柱壳求解时非常方便。算例分析的结果验证了本方法的正确性     

求解卷积公式的一种新方法

基于数形结合方法以及变量代换方法发现计算卷积公式存在某些不足之处,研究卷积公式的计算方法,给出一种新的确定积分限的方法.该方法直观明了、容易理解和掌握并且可操作性强,有效避免作图和分析图形的繁琐,并举例说明了该应用方法.     

基于随机森林的重要性测度指标体系

重要性测度分析可以找出重要特征变量,从而降低输入空间的维数,节约运算成本。基于随机森林重要性测度的分析原理,探寻随机森林的重要性测度指标与基于方差的全局灵敏度指标之间的联系,得到求解方差灵敏度主指标S_i及其总指标S_i~T的新途径。建立基于随机森林的单变量、组变量重要性测度指标,并明确具体的求解过程,完善基于随机森林的重要性测度指标体系。通过算例验证了所提基于随机森林的重要性测度指标体系的有效性及其与方差灵敏度指标之间关系的正确性。     

The incentive effect of acceptance sampling plans in a supply chain with endogenous product quality

This article studies a firm that procures a product from a supplier. The quality of each product unit is measured by a continuous variable that follows a normal distribution and is correlated within a batch. The firm conducts an inspection and pays the supplier only if the product batch passes the inspection. The inspection not only serves the purpose of preventing a bad batch from reaching customers but also offers the supplier an incentive to improve product quality. The firm determines the acceptance sampling plan, and the supplier determines the quality effort level in either a simultaneous game or a Stackelberg leadership game, in which both parties share inspection cost and recall loss caused by low product quality. In the simultaneous game, we identify the Nash equilibrium form, provide sufficient conditions that guarantee the existence of a pure strategy Nash equilibrium, and find parameter settings under which the decentralized and centralized supply chains achieve the same outcome. By numerical experiments, we show that the firm's acceptance sampling plan and the supplier's quality effort level are sensitive to both the recall loss sharing ratio and the game format (i.e., the precommitment assumption of the inspection policy). © 2013 Wiley Periodicals, Inc. Naval Research Logistics, 2013     

An MX/G/1 queueing system with disasters and repairs under a multiple adapted vacation policy

We consider a queueing system with batch Poisson arrivals subject to disasters which occur independently according to a Poisson process but affect the system only when the server is busy, in which case the system is cleared of all customers. Following a disaster that affects the system, the server initiates a repair period during which arriving customers accumulate without receiving service. The server operates under a Multiple Adapted Vacation policy. The stationary regime of this process is analyzed using the supplementary variables method. We obtain the probability generating function of the number of customers in the system, the fraction of customers who complete service, and the Laplace transform of the system time of a typical customer in stationarity. The stability condition for the system and the Laplace transform of the time between two consecutive disasters affecting the system is obtained by analyzing an embedded Markov renewal process. The statistical characteristics of the batches that complete service without being affected by disasters and those of the partially served batches are also derived. © 2015 Wiley Periodicals, Inc. Naval Research Logistics 62: 171–189, 2015     

主方差分析方法

该文提出一种对高维随机向量 X=( x1 ,x2 ,… ,xp)′p× 1 进行降维处理的实用方法 ,其基本思想是利用矩阵的扫描运算 ,构造 X的很少几个综合指标 (称为主方差变量 )以反映 X的统计特性。给出了该方法的理论依据和直观解释以及算法。特别指出 ,当变量 X多重相关性突出时 ,该文方法显著地优于主成分分析方法。