关于迭代数列的审敛法

从迭代数列及其基本性质出发,给出单调有界定理、压缩映象原理、Cauchy收敛准则和上(下)极限四种判别迭代数列收敛的方法.     

不完全语言信息下的多准则群决策方法研究

针对决策者所给的自然语言信息缺失判断矩阵,提出了一种基于群体满意度最大的不完全语言信息多准则群决策规划模型.首先分析决策者所给的多准则语言评价信息矩阵,进而通过三角模糊数将多准则语言评价信息矩阵转化为三角模糊数矩阵;其次根据满意度函数构建不完全语言信息多准则群决策规划模型;最后通过实例验证本方法的可行性及有效性.实例表明该方法计算简单,易操作.     

THE CRITERION FOR THE STABILITY OF A MULTIPLE LIMIT CYCLE UNDER HIGHER DEGENERATIONS

In this paper, we obtain the exact computation formulae to determine the stability of a multiple limit cycle with the third or fourth order degenerations. We employ the method of computing the expansion of the Poincar′e map around the closed orbit using "normal bundle" coordinates parameterized by time variable in a neighborhood of the closed orbit. An example is given to show the feasibility of our results.     

带随机初值的随机偏微分方程整体解的存在性

研究了一类带随机初值并且由分数次Brownian运动驱动的随机偏微分方程.借助于Kolmogorov准则,建立了整体Lipschitz条件下此类随机偏微分方程的一个解.同时证明了局部Lipschitz条件下整体解的存在性.     

解非线性规划问题的不精确线性搜索SQP滤子方法

本文用序列二次规划方法(SQP)结合Wolfe-Powell不精确线性搜索准则求解非线性规划问题.Wolfe-Powell准则是一种能够使目标函数获得充分下降而运行时间较省的确定步长方法.不精确线性搜索滤子方法比较其它结合精确线性搜索和信赖域方法求解问题的滤子方法更灵活更易实现.如果目标函数的预测下降量为负,我们的工作将主要利用可行恢复项改善可行性.一般条件下,本文提出的算法较易实现,且具有全局收敛性.数值试验显示了算法的有效性.     

粗信息矩阵的数量特征

利用已给出粗信息矩阵及粗集的有关理论,提出并研究了粗信息矩阵的两个数量特征--粗牲度短阵及粗精度矩阵,给出了一系列性质及定理.     

固定接触界面切向静弹性刚度问题研究

根据两球体单峰同时受法向、切向载荷时微滑切向应力的分布以及MB模型,给出了界面的总切向接触静弹性条件刚度、总条件法向载荷的解析解。将切向接触静弹性条件刚度的解析解嵌入到有限元软件中,获得了整机的理论模态。以一款八四七厂华中工学院XHK5140型自动换刀计算机数控立式镗铣床上的结合部为研究对象,通过实验对解析解进行了定量验证。研究结果表明：在理论振型与实验振型一致的条件下,界面模型的相对误差在-19．2％～16．8％之间。     

Blow-up criterion for compressible MHD equations

In this paper, we study the 3D compressible magnetohydrodynamic equations. We obtain a blow up criterion for the local strong solutions just in terms of the gradient of the velocity, similar to the Beal-Kato-Majda criterion (see J.T. Beal, T. Kato and A. Majda (1984) [1]) for the ideal incompressible flow. In addition, initial vacuum is allowed in our case.     

The impact of sharing customer returns information in a supply chain with and without a buyback policy

In this paper, we examine a single period problem in a supply chain in which a Stackelberg manufacturer supplies a product to a retailer who faces customer returns and demand uncertainty. We show that the manufacturer incurs a significant profit loss with and without a buyback policy if it fails to account for customer returns in the wholesale price decision. Under the assumption that the retailer is better informed than the manufacturer on customer returns information, we show that without a buyback policy, the retailer prefers not to share if the manufacturer overestimates while it prefers to share customer returns information if the manufacturer underestimates this information. If the manufacturer offers a buyback policy, we have the opposite results. We also discuss incentives to share the customer returns information and some of the issues that are raised in sharing this information.