ID 原文 译文
55617 实验结果表明,该算法能有较低的误分类误差,能有效地分割出车底阴影。 The experimental results show that the algorithm can have lower misclassification error, can effectively segment the car shadow.
55618 为了改善小擦地角海杂波建模存在的重拖尾问题,研究KK分布建模方法,基于L波段高分辨小擦地角海杂波数据,分析不同分辨率和不同海况下KK分布与海杂波实测数据的拟合效果,通过与常用的K分布、威布尔分布、对数正态分布等对比,表明KK分布较好解决了K分布等在幅度分布拖尾处与实测数据出现的偏离问题,可在拖尾处达到更好的拟合效果。 In order to improve the small brush seed weight trailing problems of sea clutter modeling, research KK distribution modeling method, based on the l-band high resolution small seed of sea clutter data, analysis of different resolutions and KK distribution under different sea conditions and sea clutter measurement data fitting effect, with the commonly used K distribution and weibull distribution, logarithmic normal distribution, such as contrast, show that KK is better solved the K distribution in amplitude distribution, such as the tail in the deviation problem with experimental data and can be at the tail reach better fitting effect.
55619 针对一类具有未知时变时滞的不确定高阶非线性系统,基于增加幂次积分方法,提出了一种非光滑状态反馈自适应神经网络动态面控制设计方案。 For a class of unknown time-varying time-delay uncertain high-order nonlinear systems, based on the increase the exponential integral method, this paper proposes a nonsmooth state feedback adaptive neural network dynamic surface control design.
55620 通过构造适当的Lyapunov-Krasovskii泛函处理了未知时变时滞不确定项;通过利用神经网络权值范数的适当形式幂次函数,将神经网络用于对在单步递推中所构造的未知函数进行建模;采用动态面技术,解决了"微分爆炸"问题。 By constructing proper Lyapunov - Krasovskii functional dealt with unknown time-varying time-delay uncertainties;By using the neural network weight norm, the exponential function with the correct form of will in stepping recursive neural network is used to unknown functions of the constructed model;By using dynamic surface technology, can solve the problem of "differential explosion".
55621 所提控制方案能够保证闭环控制系统的状态量和跟踪误差半全局一致终结有界。 The proposed control scheme can guarantee the closed-loop control system of the state and global asymptotical end half tracing error to be bounded.
55622 最后,仿真算例结果表明了该方案的有效性。 Finally, the simulation example results show the effectiveness of the scheme.
55623 研究了在一个固定区域对随机运动目标进行连续探测的概率时间收益问题。 Studied in a fixed area for continuous random motion target detection probability time benefit.
55624 首先建立了"绝对连续探测"的概念,证明了在绝对连续探测条件下,在一个区域的持续探测,持续时间不会带来概率的增加。 First establish the concept of "absolutely continuous detection", proved that under the condition of absolutely continuous exploration, in an area of continuous detection, duration will not bring the increase of the probability.
55625 提出了"一次探测周期"概念,建立了一定"一次探测周期"时持续探测的概率与周期数的函数关系,并定义了持续探测概率的概率收益、概率时间收益率。 Proposed the concept of "a detection cycle", established a certain "a detection period" for the probability of detection as a function of cycle number, and defined the probability of detection probability and the probability of the earnings, return on time.
55626 算例表明了所定义各参数的基本变化形式。 Numerical examples show that the basic change of the parameters are defined.