对大词汇量汉语连续语音识别的实验结果表明：高斯模糊聚类使高斯数减少25%时，识别率提高了0.15%。
The experimental results on large vocabulary continuous Mandarin speech recognition show when the number of Gaussians is reduced by 25%, the recognition accuracy increases by 0.15%.
偿试利用二元光学的方法选择合适数的相位片对光束进行整形，并就高斯光束均匀化为例进行了数值计算。
In the paper, it is attempted to choose suitable slices of phase for beam shaping by use of binary optics, for example, the uniformization of Gaussian beam is calculated numerically.
用高斯-牛顿误差最小法将六维观测量转化为四元数，作为观测量的一部分，显著减少了直接使用EKF的计算量。
Gauss-Newton error minimization is used to transform six-dimentional reference vector to quaternion as a part of observations for EKF, which significantly reduces the computational requirement.
本文提出了一种生成广义高斯分布(GGD)随机数的通用算法。
In this paper, a new algorithm is proposed to generate generalized Gaussian distribution (GGD).
依据中心极限定理，用均匀分布随机数求和的方法得到趋于高斯分布的白噪声。
White noise tending to Gaussian distribution is implemented by summing uniformly distributed random numbers according to the central limit theorem.
基于高斯摄动方程，推导了卫星在同时考虑J2和大气摄动情况下的轨道根数变化方程。
The variation equation of the satellite orbit elements under both J2 and drag perturbations was derived based on the Gauss perturbation equation.
设计了一个基于FPGA的高速、高性能的高斯随机数发生器。
A highspeed gaussian random number generator implemented with FPGA is presented in this paper.
通过对四分位数稳健统计进行z比分数法能力验证过程的分析提出高斯节点z比分数法。
By quartile robust statistical Z-Score way analysis of Proficiency Testing process to make Gaussian node Z-scores ways.
对给定调制参数的余弦-高斯光束，通过适当选取分数傅里叶变化阶数可以获得平顶的光强分布。
A flat-topped beam can be achieved by a suitable choice of the transform order for the Cosine-Gaussian beams with the modulation parameter.
还研究了超高斯光束通过近轴光学系统时，保持其形状和阶数不变的条件。
The condition has been discussed, under which SG beams preserve their shape and order, while passing through paraxial optical systems.
为设计一个用于信道仿真和编码器性能测试的高速高斯随机数发生器，研究了适于硬件实现的高速通用的连续随机变量和随机序列产生法。
To design a gaussian random number generator used for channel simulation and coder performance test, high-speed universal random number and correlated stochastic series generators are studied.
数值计算表明，聚焦双曲正弦高斯光束的焦移与偏心参数和菲涅耳数有关，焦移随偏心参数和菲涅耳数的减小而增大。
Numerical calculations are performed and show that the focal shift in focused sinhGaussian beams depends on the Fresnel number and decentered parameter, and increases with decreasing them.
由高斯取整的性质和集合的无重复性，此集合的基数必然是一个小于200的未知数。
The radix of this set must be a unknown number smaller than 200 because of the characters of Gauss function and the repeatlessness of set.
最后，把菲涅耳数推广到复数域，并用于研究截断高斯光束的衍射问题。
Finally we extend the Fresnel number to the region of complex number, and use it to investigate the diffraction problem of a truncated Gaussian beam.
在用蒙特卡罗法进行仿真研究(例如进行测量不确定度评定)时，常常需要发生多个非高斯型互相关的随机数。
In Monte Carlo simulation such as the measurement uncertainty evaluation, it is often necessary to generate correlated multi-non-Gaussian random observations.
以伪均匀随机数为基础，根据大数中心极限定理，产生高斯分布随机数。
Based on pseudo random uniformity number and the central limit theory of great numeral, random Numbers of Gaussian distribution are produced.
本文分析高斯光束菲涅耳数和聚焦高斯光束各参数之间的关系，给出了一种简便的设计方法。
The relations between the Fresnel number of a Gaussian beam and the parameters of the focusing Gaussian beam are analysed in this paper. And a simple method of design is given.
结果表明：对理想超高斯光束，半高全宽内包含的能量百分数最大不超过70%，且随着阶数增大而逐渐降低。
The results show that the percentage of energy in FWHM of the ideal super-Gaussion beam is no more than 70%, and it reduces with the order of the super-Gaussian beam increasing.
结果表明：对理想超高斯光束，半高全宽内包含的能量百分数最大不超过70%，且随着阶数增大而逐渐降低。
The results show that the percentage of energy in FWHM of the ideal super-Gaussion beam is no more than 70%, and it reduces with the order of the super-Gaussian beam increasing.
应用推荐