ID 原文 译文
5564 针对现有航迹起始方法难以对编队目标进行有效航迹起始的问题,在Hough变换法及其衍生算法基础上,提出基于Hough变换和高斯混合最大期望(expactation maximazation,EM)聚类的多编队目标航迹起始方法。 In view of the existing track initiation methods effectively for the groups are hard to track the initial problem, based on Hough transform method and its derivative algorithm, is put forward based on Hough transform and gaussian mixture maximum expected (expactation maximazation, EM) cluster formation of target track initiation method.
5565 该方法首先利用量测数据的时序信息和目标的运动参数进行筛选,剔除大量虚假量测;再对筛选后的量测数据进行Hough变换,得到初步航迹信息;然后利用相异度矩阵对所得航迹进行预聚类,完成聚类中心初始化;最后进行高斯混合EM聚类,得到聚类结果。 This method firstly use the test data of temporal information filtered and target motion parameters, eliminate a large number of false measurement;Again after the screening of Hough transform of the measuring data and get preliminary tracking information;Cyber matrix is then used to preliminary clustering of the track, complete initialization clustering center;Finally gaussian mixture EM clustering, clustering results are obtained.
5566 仿真结果表明,与Hough变换法及其衍生算法相比,该方法能够快速有效地起始编队目标的航迹,解决了目标密集带来的航迹起始混乱问题。 The simulation results show that compared with Hough transform method and its derivative algorithm, this method can quickly and efficiently starting formation of target track, solves the target track is initial chaos of the problems of intensive.
5567 为了提高从宽角合成孔径雷达(synthetic aperture radar,SAR)图像中提取目标后向散射各向异性特性的性能,在宽角SAR字典稀疏表示模型的基础上,提出一种基于高斯字典原子的高精度宽角SAR成像方法。 In order to improve the light Angle of synthetic aperture radar (synthetic aperture radar, SAR) image to extract target to the scattering of anisotropic characteristics of performance, in the wide Angle SAR dictionary based on sparse representation model, put forward a kind of high precision based on gaussian dictionary atoms wide Angle SAR imaging method.
5568 在字典构造上,采用不同中心位置、相同方差的高斯函数。 In the structure of the dictionary, use different center position, the same variance of gaussian function.
5569 在求解稀疏表示系数上,采用广义最小最大凹惩罚稀疏重构算法求解。 On the coefficient of solving sparse representation, by using the generalized minimum concave biggest punishment sparse reconstruction algorithm.
5570 最后,根据稀疏表示系数的重构结果以及构造的字典得到目标的后向散射各向异性特性。 Finally, according to the results of reconstruction sparse representation coefficient and structural dictionary backscatter anisotropic characteristics of the target is obtained.
5571 通过仿真实验和Backhoe数据对算法进行验证,结果表明,该方法能够高精度地提取目标的后向散射各向异性特性。 By simulation experiment and Backhoe data to verify this algorithm, the results show that the method can accurately extract the target backscatter anisotropic characteristics.
5572 针对切换拓扑结构下的集群编队控制问题,设计只需个别无人机获取虚拟长机信息也能保证集群连通性的编队控制算法。 For cluster formation control problem of switching topology, design simply individual unmanned aerial vehicle (uav) for virtual flight information also can assure the connectedness cluster formation control algorithm.
5573 当队形变换或部分通讯网络故障导致网络拓扑结构发生改变时,以距离为原则对集群进行联盟划分,各联盟内部成员以信息浓度大小为标准同其他成员进行竞争,由信息浓度最大的无人机获取虚拟长机信息,其联盟成员通过与该无人机通讯间接获取虚拟长机信息。 When formation transformation or part of a communication network failure network topology changes, based on the principle of cluster alliance distance, all the members of the alliance with the size as the standard information concentration to compete with other members, by the information concentration's largest unmanned aerial vehicle (uav) for virtual flight information, its alliance with the uav communication indirectly get the virtual flight information.