ID	原文	译文
6204	然后用加权最小二乘法给出了寿命分布的参数估计,进而得到了装备贮存可靠度的估计和表达式。	Then the life distribution is solved using the weighted least squares parameter estimation, then get the equipment storage reliability estimation and expression.
6205	最后结合实例进行计算,说明了方法的稳定性和可行性,为工程上评估装备贮存可靠度提供借鉴。	Finally calculated based on an example, the stability and feasibility of the method is shown for engineering evaluation equipment of storage reliability for reference.
6206	针对异常数据和/或数据序列的检测,根据再生核希尔伯特空间最大平均偏差异常数据和/或数据序列检测算法,发展出了一种恒虚警检测异常的非参数方法。	For abnormal data and/or data sequence detection, based on reproducing kernel Hilbert space maximum deviation from the mean abnormal data and/or data sequence detection algorithm, developed a method to detect abnormal nonparametric constant false alarm.
6207	将来自正常数据的最大平均偏差描述成了一个统计分布。	Described the maximum average deviation from the normal data into a statistical distribution.
6208	分析表明:奈曼—皮尔逊假设检验可利用这个分布来进行异常假设检验,而bootstrap重采样技术或期望最大算法则可估计出正常数据或数据序列的统计分布。	Analysis shows that nai man - Pearson hypothesis test available for abnormal hypothesis test, the distribution and the bootstrap resampling technique or expectation maximum algorithm can estimate the normal data or statistical distribution of data sequence.
6209	尽管在给定虚警率的条件下,异常假设检验所需的判决门限可由估计到的统计分布计算获得,但可以利用蒙特卡罗积分的方法来简化这个计算。	Although under the condition of a given false alarm rate, abnormal decision threshold needed for hypothesis testing can be calculated by the estimated to the statistical distribution of gain, but can use monte carlo integral method to simplify the calculation.
6210	数值仿真的结果验证了提出方法的有效性,同时,表明所提方法优于文献中报道的方法。	Numerical simulation results verify the effectiveness of the proposed method, at the same time, show that the proposed method is superior to the method reported in the literature.
6211	针对传统的核相关滤波跟踪算法特征表达不足且不能解决尺度变化导致的跟踪失败问题,提出了一种多特征融合的尺度自适应相关滤波跟踪算法。	Expression based on the traditional nuclear related filter tracking algorithm is insufficient and can not solve the problem of scale changes caused by tracking failure, this paper proposes a multiple characteristics of the fusion of the scale of the adaptive filter tracking algorithm.
6212	首先,在位置预测阶段,利用Hue和梯度方向直方图两种特征和各自的位置滤波器作用得到的输出响应值,自适应分配权重实现目标位置的预测。	First of all, in the location prediction stage, the use of two Hue and gradient direction histogram features and their respective location filter function to get the output response value, adaptive allocation weights to achieve target location prediction.
6213	其次,在尺度预测阶段,通过在目标位置采集多尺度图像,组成样本独立训练尺度滤波器,并根据样本的尺度滤波响应值,对目标的尺度进行估计,使跟踪算法能够适应目标的尺度变化。	Second, the scale prediction stage, through multi-scale image acquisition in the target position, independent of sample training scale filter, and according to the filter response values of the scale of the samples, to estimate the scale of the target, make tracking algorithm can adapt to the change of the scale of the target.