弱收敛

In order to get the convergence properties of the weak set-valued Amart , we firstly proved the theorem that the limit of support functions is a support function.

为了得到关于弱集值渐近鞅的收敛性质，首先证明了支撑函数列的极限亦为一支撑函数。

A proper k total colouring of a graph G is a colouring to its vertices and edges using k colours such that no two adjacent or incident elements of G may be assigned the same colour.

利用梯度投影与罚函数相结合的技巧，将带不等式和等式约束的优化问题化成一个无约束问题，提出了初始点可任意的求解不等式、等式约束优化问题的摄动梯度投影算法；参数δk取不同的数还可以得到一类梯度投影算法，从而得出了在搜索方向和步长不精确条件下的梯度投影法，保证了在实际应用中更容易实现；在较弱条件下，证明了该算法的全局收敛性。

Loosedthe illter-elemefit colltinuity further with the requirement of Irons patch teSt,obtains the weak continuity condition ensured convergence.

用分片试验的要求来进一步放松单元间的连续性，可以得到保证收敛的弱连续条件，它只包括单元顶点的函数值连续和法向导数积分连续2个条件，从而扩大了有限元的构造范围。

Modify the non-conformin8 nominal derivation of BCIZ element accordin8 tothe weak continuity conditiory obtained convergent element too.

考察弱连续条件下的有限元，结果它们具有相同的平均应变，其它收敛的有限元、如离散的克希霍夫板元、也具有这种平均应变。

Loosedthe inter-element continuity further with the requirement of Irons'patch test,obtains the weak continuity condition ensured convergence.It comprises thefunction value at the node continue and its nominal derivative continue 2conditions,hence do enlarge the constructing area of the finite element.

用分片试验的要求来进一步放松单元间的连续性，可以得到保证收敛的弱连续条件，它只包括单元顶点的函数值连续和法向导数积分连续2个条件，从而扩大了有限元的构造范围。

But the active contour model has some disadvantages such as the high complexity of the algorithm and its poor convergence to the weak edges of images.

但是，活动轮廓模型本身也存在很多缺陷，如计算复杂度比较高、图像的弱边缘收敛性比较差等问题。

According to the characteristics of strong global search, but weak local search and slow convergence of genetic algorithm, this paper proposed a new hybrid algorithm which combined GA and PSO, and applied it successfully to automatic software test data generation.

针对遗传算法具有较强的全局搜索能力，但局部搜索能力较弱，且收敛速度慢的特点。将遗传算法与粒子群算法结合起来形成新的混合算法，并成功应用到软件测试数据自动生成过程中。

An improved real-code genetic algorithm is designed with different cross, selection and variation operators, aiming at overcoming its inherent shortcomings of premature, dead state convergence and weak mountain climbing during function optimization.

针对实数编码在进行数值优化时固有的过早收敛、停滞现象和弱的爬山能力等缺点，通过设计不同的交叉、变异算子，提出了一种改进的实数编码遗传算法。

But the weak law of large numbers and the complete convergence for arrays of rowwiseρ*-mixing random variables have not been reported.

但是关于它所生成的随机变量组列的弱大数定律及完全收敛性还未见报道，本章讨论这方面的内容，仅在条件ρ~*(1)＜1，而对混合速度不作任何限制下，得到了：定理0.1.1设{X_；1≤k≤n，n≥1}是行内随机变量为ρ~*混合且ρ~*(1)＜1的随机变量组列。

There are some disadvantages in BP algorithm, such as low rate of convergence, easily falling into local minimum point and weak global search capability. In order to settle these problems, a genetic algorithm was used to train BP neural network to replace classical learning algorithms. An evolutionary neural network learning algorithm was founded.

针对BP算法存在收敛速度慢、容易陷入局部极小值及全局搜索能力弱等缺陷，采用遗传算法训练BP神经网络，取代了一些传统的学习算法，设计了基于进化神经网络的学习算法。

Chrysanthemum of 10 thousand birthday is lax to edaphic requirement, with the arenaceous qualitative loam with fecund, good drainage had better.

万寿菊对土壤要求不严，以肥沃。排水良好的砂质壤土为好。

He unstepped the mast and furled the sail and tied it.

他拔下桅杆，把帆卷起，系住。