查询词典 numerical algorithm

With the use of analytical and numerical analysis to analyse the stability and attractability of the network using the two learning rules, the associative memory of the network using higher-order projection algorithm is better than that of using higher-order Hebbian algorithm.

通过在解析和数值两方面对这两种学习规则下网络的稳定性和吸引性的分析，可以看出高阶投影规则下网络的联想记忆性能比高阶Hebb规则下的要好。

Based on the harmonic algorithm, the conductivity reconstruction method and the corresponding algorithm in oblique slice were proposed. Then the numerical simulation was performed by solving the Neumann boundary problem, Ohm's law and Biot-Savart law to obtain the magnetic flux density used to reconstruct the conductivity.

然后通过数值仿真来实现提出的算法，首先进行了正问题求解，即利用Neumann边界值问题、欧姆定律和毕奥－萨伐尔定律逐步计算出重建电导率分布的磁通密度的一个分量，接着进行了逆问题求解，即根据重建算法得到了横截面和斜位面内的相对电导率分布，并对仿真的结果进行了分析和讨论。

The numerical experiments shows that the algorithm is much more efficient than the general genetic algorithm and conjugate gradient method.

算例结果表明，该算法对于病态方程组的求解效果明显优于一般的遗传算法和共轭梯度法。

In this paper, we propose an algorithm for finding global minimization by discrete mean value-level set, the convergence of this algorithm is proved, a termination rule and some numerical examples are given.

西文提出了一个只需计算函数值的离散均值—水平集求全局最优的方法，给出了算法的终止准则，证明了算法的收敛性，并给出了某些数值例子。

Numerical experiments show that the new algorithm is more effective in realizing the high convergence speed, convergence precision, reducing the convergence algebra and good at keeping the stability of the adaptive genetic algorithm.

数值实验表明：新算法在提高收敛速度和收敛精度、减少收敛代数方面效果显著，稳定性也有所提高。

The paper taked 20 nodes isoparametric element as an example and put forward a kind of numerical method combined with the Newton iterative algorithm and Gauss integral algorithm to calculate equivalent loads of prestress in body element.

本文以20节点等参单元为例，提出一种运用高斯积分和牛顿迭代法的数值算法计算实体单元的预应力等代荷载，该方法能够适应各种复杂的预应力钢束形状，并适合计算机编程。

Numerical experiments show that the updating algorithm by using intergeneration differential and chaotic mutation (algorithm 8) is a good selection. Simultaneously, the updating effect is perfect when the updated particles are about 20%.

通过数值实验得出基于代间差分和混沌变异的粒子更新算法（即算法8）是一种很好的选择，并且当性能较差的20%左右粒子按照这种算法更新时效果较好。

With the application of the sampling theorem and the Fast Fourier Transform, two groups of numerical forms of the algorithm equations suitable for two kinds of FFT subroutines are derived. Finally, the measurement algorithm is fully tested by the measurement and calculation carrier out for an MMW irradiator.

在解析算法的基础上，应用采样定理和快速傅里叶变换，推导了分别适用于两种快速傅里叶变换子程序的数值算法公式，通过对一种毫米波辐射器的实际测量和计算对测量算法进行了整体检验。

The main results of this research are as follows:(1) A parallel algorithm for solving dense matrix linear equations on distributed-memory multi-computer is presented. Convergent velocity of this algorithm is better than parallel Jacobi method"s, and parallelism is better than Gauss-Seidel method"s. According to theoretical analysis, convergent conditions are given. Moreover numerical experiments indicate that results agree with theory.

主要完成了如下研究工作：(1)提出了一种求解系数矩阵为稠密矩阵的线性方程组的适合于MIMD分布式存储的并行迭代算法，该算法设计上兼取了Jacobi算法并行性好的优点和Gauss-Seidel算法收敛速度快的长处，并且在理论上给出了该算法成立的收敛性条件，并通过数值计算验证了此算法结果与理论相符合。

Comparing with traditional LU numerical decomposition algorithm to solve each row of lower triangular matrix L and each column of upper triangular matrix U, the proposed method is easy to program and efficient In addition, the structure symmetry of Jacobian matrix and the experience of compiler optimization are also adopted in this research, it makes less required memory for the proposed algorithm and higher calculation efficiency.

这种算法和求取L阵每列和U阵每行的传统LU分解方法相比，具有编程简单、计算效率高的优点。另外，雅可比矩阵结构对称以及编译器优化的经验也应用到文中，使得算法不仅占用内存较少，且效率较高。

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。