算子函数

In this paper,a systematic direct perturbation method of dark solitons is found.Having analyzed the mistakes in earlier works on perturbation method for dark solitonsand essence of the direct perturbation method for bright solitons,we notice that to in-troduce the adjoint solutions of the squared Jost solutions and to prove the completenessare crucial to the problem.Giving up the unnecessary scheme of introducing the adjointoperator in the bright soliton case,we directly find the adjoint solutions by meetingthe demand for the orthogonality that inner product of the squared Jost solutions andits adjoint should be proportional to a δ function in the case of continuous spectra.The corresponding adjoint operator is thus found.Taking into account the reductiontransformation,we find a correct description for the completeness of the squared Jostsolutions and directly verify its validity with explicit expressions of the squared Jostsolutions.

本论文建立了系统的暗孤子直接微扰方法，在对前人关于暗孤子微扰方法的错误以及亮孤子直接微扰方法的本质作了充分的分析后，认识到引入平方Jost解的伴随解和证明完备性是问题的关键，撇开过去亮孤子情况首先引入伴随算子的非必要作法，直接从平方Jost解与其伴随解的内积在连续谱时正比于δ函数这一正交性要求出发，找出了伴随解，同时得出了应有的伴随算子，在考虑到约化变换性后，得到了暗孤子情况的平方Jost解的完备性的正确表述，并在单个暗孤子的情况利用平方Jost解的显式直接验证了它的正确性。

In Chapter Three, we use the method of interpolation spline of differential operater to come up with the reproducing kernel in H01 with respect to bounded linear operator in H10. Then we use the reproducing kernel to develop the expression of the best approximating of bounded linear operator in H10 and prove its convergence.

第三章中，对于H_0~1中的有界线性算子，用微分算子插值样条函数的方法给出了H_0~1空间中的再生核，利用此再生核给出了H_0~1上的有界线性算子的最佳逼近的表达形式，并证明了其收敛性。

Firstly,we introduce its definition,then construct two groups of operator functions by an elementary method,and extend the function f= t(t+2)log(t+2)/(t+1)~2 to linear situation,at last we discuss some results about operator monotone function under the complex order and conjugate complex order.

首先介绍了Furuta不等式的几何结构，将普通的算子序推广到更弱的算子序A~4≥(A~2B~2A~2)~2/3，讨论在此序下的函数的算子单调性，并且推广到更一般的序A~(2m)≥(A~mB~2A~m)~m/m+1，得到了一些新的研究方法与结果。

Secondly, a projector operator Q is defined in the Lebesgue integrable functions' space Y such that the composition of the inverse operator of L, I-Q and the nonlinear term f is compact, where I is the identity mapping in Y.

然后，在Lebesgue可积函数全体所成的空间Y上定义了投影算子Q，使得L的逆与I-Q及非线性项f的复合是紧算子，其中，I是Y上的恒同算子。

The following results are presented: By the property of mean value of the holomorphic functions and the property of the reproducing kernel k_ω, we get the rank one operator k_ω?

首先，利用正规化再生核K_w的特殊性和解析函数的平均值性质，建立了秩为一的算子K_wK_w与Toeplitz算子之间的联系，把算子K_w？

This method was mainly built on the operator matrix's singular value decomposition in the first class of operator equations, and then the best regularization parameters were selected in accordance with the characteristics of singular value so as to identify filter function, thereby removing noise of the function.

该方法主要建立在第一类算子方程的算子矩阵进行奇异值分解的基础上，然后根据奇异值的特征选取最佳正则化参数确定滤子函数，从而达到对该类函数的噪声去除效果。

The core designs of genetic algorithm are design of parameter coding, initial group, adaptitude function, controlling parameter and genetic operators. On other words, they are designs of choose operators, exchange operators and aberrance operators.

遗传算法的核心是参数编码、初始群体的设定、适应度函数的设计、控制参数设定、遗传算子的设计即选择算子、交换算子和变异算子的设计。

Firstly, from the functional analysis, the eigenfunction series of vectorial partial differential operator is studied. The property of complete space is gotten. And this result is extended to the eigenfunction of Maxwells equations.

本文首先从泛函分析的方法出发，研究了矢量偏微分算子本征函数系的性质，得到了矢量偏微分算子正交完备归一化本征函数的存在证明，研究了Maxwell方程组的情况，证明了电磁场在均匀和非均匀介质条件下的正交本征函数系的完备性质。

An improved TennenGrad criterion function and an improved Laplace criterion function are proposed for microscopic coarse autofocusing, which use Sobel operators and Laplacian operators to convolve and add together the absolute values of matrix pixels for a image.

提出了使用Sobel算子或Laplace算子做卷积，对图像像素点进行绝对值相加的TennenGrad改进算法，以及La-place改进算法，和几种已有评价函数进行分析比较。为了判断粗调评价函数的优劣，对评价函数的判断准则进行了逐一分析，并且将平滑性、高效率、强壮性作为最主要的衡量标准。

The results of using three kinds of aggregator with parameter-Dombi aggregator, Yager aggregator and Mean aggregator are compared.

并采用了三种典型的带参数的合成算子：Yager T算子，Dombi T算子和广义Mean算子，合成模糊决策目标函数。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。