ID 原文 译文
38506 该方法利用最小二乘法拟合算法解决移相器只能提供离散相位变化,不能准确获得使天线合成功率最大时的移相器相位变化值的问题。 The method using a least squares fitting method to solve the phase shifter is able to provide discrete phase change, can not accurately obtain the antenna phase shifter for synthesis of the largest power values of the phase change problem.
38507 通过构造伪Hadamard矩阵对天线分组,每一组移相器同时旋转多个阵元相位,使被测信号变化明显。 By constructing pseudo Hadamard matrix for antenna group, each group phase shifter rotate multiple array yuan phase at the same time, make the measured signal change is obvious.
38508 远场校准实验表明,该方法能够对弹载相控阵天线起到非常好的校准效果。 Far field calibration experiments show that this method is able to onboard calibration phased array antenna have very good effect.
38509 将高维特征用于跳频电台细微特征个体识别具有很大优势,为了增强对跳频电台的分类识别能力,需要增加特征类型和维数,提高特征集的表征能力,但同时会引入大量冗余特征,导致分类器计算时间过长,分类正确率降低。 The high-dimensional features for fh radio fine individual identification has a big advantage, in order to enhance the classification ability of frequency hopping radio, need to increase the characteristics of the type and dimension, and improve the representation ability of feature set, but at the same time will introduce a large number of redundant features, computation time is too long, cause classifier classification accuracy is reduced.
38510 为了降低高维特征集维数,首先采用相关性快速过滤特征选择算法,删除高维特征集中的不相关冗余特征,得到最优特征集。 To reduce high victor solicitation dimension, first USES the correlation filter feature selection algorithms, quickly remove high-dimensional feature concentration of unrelated redundant features, its collection.
38511 然后利用经过参数优化的支持向量机(support vector machine,SVM)分类器进行训练分类。 Then using the optimized parameters of support vector machine (support vector machine, SVM) classifier was trained classification.
38512 实验表明,所提算法能够对高维特征集进行合理的降维,提高了SVM的分类器的分类性能,在保证分类正确率的基础上,降低了运算量,提高了跳频电台细微特征识别的时效性。 Experiments show that the proposed algorithm is able to carry on the reasonable dimension reduction of high victor solicitation, improves the classification performance of the SVM classifier, based on the classification accuracy, reduces the computational complexity, improves the efficiency of the frequency hopping radio fine feature recognition.
38513 软件可靠性增长模型一般假设故障是独立的,且检测到的故障能够被排除。 Software reliability growth models generally assume that the fault is independent, and the detected fault can be ruled out.
38514 但在工程中,一些检测到的故障可能无法排除,在排除过程中也可能会引入新的故障。 But in the engineering, some of the detected fault may not be able to rule out, also in the process of rule out may introduce new fault.
38515 考虑上述因素,本文假设故障的引入过程与时间呈现非线性关系,通过测试覆盖率函数来表示软件故障检测率,建立了基于非齐次泊松过程(non-homogeneous Poisson process,NHPP)的软件可靠性增长模型。 Considering the above factors, this article assumes that the introduction of the failure process of a non-linear relation with time, through the test coverage function to represent the software fault detection rate, based on the nonhomogeneous Poisson process (non - homogeneous Poisson process, NHPP) software reliability growth model.