ID 原文 译文
11814 并在此基础上,建立了综合考虑这两种效用的软件可靠性增长模型。 And on this basis, established a comprehensive consideration of the two utility software reliability growth model.
11815 利用公开发表的软件测试失效数据集对模型进行验证。 Use of published software test failure data sets to verify this model.
11816 结果表明,与现有的考虑故障检测与修正过程的模型相比,所提模型具有更好的拟合和预测能力。 The results show that with the existing consider compared to fault detection and correction process model, the proposed model has better fitting and forecasting ability.
11817 压缩感知理论中信号的重建要求量测矩阵与稀疏变换基之间的互相关性要尽可能小。 The reconstruction of the signal in the compressed sensing theory requires that the measurement matrix and sparse cross correlation between the base should be as small as possible.
11818 以降低二者互相关性为目的,研究了一种改进的基于变步长梯度下降的量测矩阵优化方法。 ‭In order to reduce the mutual correlation for the purpose, research a kind of improved based on variable step length measurement matrix optimization method of gradient descent.
11819 该方法利用梯度下降法更新步长,并基于模拟退火中的降温思想引入学习速率因子来进一步调节步长的变化,提高算法的收敛速度,改善算法的性能。 ‭The update step length method using the gradient descent method, and based on simulated annealing cooling of the learning rate factor are introduced to further adjust the step change, improve the rate of convergence of the algorithm, improves the performance of the algorithm.
11820 仿真结果表明,使用变步长梯度下降法优化后的量测矩阵与稀疏变换基的互相关系数在零附近的分布更加集中,量测矩阵的优化速度快并且重构图像的峰值信噪比提高。 ‭The simulation results show that using variable step length of the gradient descent method to optimize measurement matrix with sparse transformation based on the number of relations between each other more concentrated distribution of near zero, the optimization of the speed measurement matrix and peak signal to noise ratio of the reconstructed image.
11821 因此,所提方法优化所得的量测矩阵无论是降低互相关性还是提高图像重建质量都具有良好的性能。 ‭Therefore, the proposed method to optimize the measurement matrix both in reducing cross correlation and improve the quality of image reconstruction with good performance. ‭
11822 在计算过程中,标准布谷鸟算法(cuckoo search algorithm,CS)中的参数是保持不变的,这影响了该算法的收敛性和计算精度。 In the process of calculation, the standard cuckoo algorithm (cuckoo search algorithm, CS) in the parameter is unchanged, this affects the convergence of the algorithm and the calculation accuracy.
11823 为了克服这一缺陷,首先探讨了标准CS中飞行步长和淘汰概率两个关键参数的变化规律对该算法全局搜索与局部搜索能力的影响,然后对这两个参数进行了自适应改进。 In order to overcome the defects, discusses the first standard in flight CS step length and elimination probability the changing rule of the two key parameters of the algorithm is global search and local search ability, and these two parameters for the adaptive improvement.