在一个新步长规则下梯度投影算法的全局收敛性

考虑约束最优化问题：minx∈Ωf(x)其中:f:R^n→R是连续可微函数，Ω是一闭凸集。本文研究了解决此问题的梯度投影方法，在步长的选取时采用了一种新的策略，在较弱的条件下，证明了梯度投影响方法的全局收敛性。     

Application of Hilbert＇s Projective Metric on Symmetric Cones

Let Ω be a symmetric cone. In this note, we introduce Hilbert＇s projective metric on Ω in terms of Jordan algebras and we apply it to prove that, given a linear invertible transformation g such that g（Ω） = Ω and a real number p, ｜p｜ 〉 1, there exists a unique element x ∈ Ω satisfying g（x） = x^p.     

一个线性约束优化问题的广义梯度投影法

本文对线性约束优化问题提出了一个新的广义梯度投影法，该算法采用了非精确线性搜索，并在每次迭代运算中结合了广义投影矩阵和变尺度方法的思想确定其搜索方向．在通常的假设条件下，证明了该算法的整体收敛性和超线性收敛速度．     

链环的trip矩阵

ZULLI L首先构造了一个用于计算纽结Kauffman尖括号多项式的模2矩阵,纽结的trip矩阵.为了构造链环的trip矩阵,引入了一个带标识的穿有m个孔的圆盘来取代纽结情形下的圆盘,其中m为链环的分支数.主要结果为:定理若状态S是从状态AA…A经过i1,i2,…,ip位置上的标记替换(A换成B)而得的状态.设Ts是将trip矩阵T的左上角的n×n子块中ai1i1,ai2i2,…,aipip之值进行替换(0→1或1→0)所得的矩阵,则#(L|S)=n+m-秩(Ts).因此计算链环Kauffman尖括号多项式就归结为计算一组模2矩阵的秩.     

关于一类非散度型非线性退化抛物方程的一些结果

In this paper, we are concerned with the initial and boundary value problem for the following equations not in divergence form:     

线性均衡约束最优化的一个广义投影强次可行方向法

本文讨论带线性均衡约束最优化问题,首先利用摄动技术和一个互补函数将问题等价转化为一般约束最优化问题,然后结合广义投影技术和强次可行方向法思想,建立了问题的一个新算法.算法在迭代过程中保证搜索方向不为零,从而使得每次迭代只需计算一次广义投影.在适当的条件下,证明了算法的全局收敛性,并对算法进行了初步的数值试验.     

自发发射与量子计算机存储单元的相干脱散

量子计算机存储单元的相干脱散，破坏量子态中的信息，是量子计算机难以实现的主要原因之一．本文在Rzazewski的工作基础上，针对两能级原子构造的存储单元，研究自发发射对存储单元相干脱散的影响 关键词：     

THE RELATIONSHIP BETWEEN PROJECTIVE GEOMETRIC AND RATIONAL QUADRATIC B-SPLINE CURVES

For two rational quadratic B-spline curves with same control vertexes, the cross ratio of four eollinear points are represented; which are any one of the vertexes, and the two points that the ray initialing from the vertex intersects with the corresponding segments of the twocurves, and the point the ray intersecting with the connecting line between the two neighboring vertexes. Different from rational quadratic Beeier curves, the value is generally related with the loeation of the ray, and the necessary and sufficient condition o5 the ratio being independent of the ray‘s loeation is showed. Alsn another cross ratio o5 the following four collinear points are suggested, i.e. one vertex, the points that the ray from the initlal vertex intersects respectivdy with the curve segmentt the line connecting the segments end points, and the line connecting the two neighboring vertexes. This cross ratio is concerned only whh the ray‘s location, butnot with the weights of the curve. Furthermore, the cross ratio is projective invariant under the projective transformation between the two segments.