ID	原文	译文
10924	提出了一种基于多任务学习的方向图可重构稀疏阵列天线设计方法。	Put forward a kind of based on multitask learning pattern reconfigurable sparse array antenna design method.
10925	将稀疏阵列优化设计及其方向图综合问题转换成为稀疏矩阵的线性回归问题,利用多任务学习能同时对多个相关任务优化学习的特性,建立了多个方向图联合赋形的多任务学习模型。	Sparse array optimization design and the conversion pattern synthesis problems become sparse matrix linear regression problems, use of multitasking can be related to multiple tasks at the same time to optimize learning features, set up a multiple pattern combined with vehicle multitasking learning model.
10926	通过迭代收缩阈值的方法,对多任务学习问题进行优化求解,使得阵列天线能够使用更少的阵元实现多个方向图的重构。	Through iterative threshold shrinkage method, optimize the multitasking learning problems to solve, the array antenna arrays can use fewer implement multiple pattern of refactoring.
10927	仿真结果表明,该方法可以生成相同阵列结构的稀布天线阵,并通过动态改变其权值向量,实现多个方向图的精确赋形。	The simulation results show that this method can generate the same antenna array, the array structure of thin cloth and by dynamically changing the weight vector, implement multiple direction figure accurate informs.
10928	为了避免着陆器在着陆过程中落于地势平坦的封闭环境,本文提出了一种基于单幅图像的星表着陆区选择方法。	In order to avoid the lander lagged landed in flat closed environment, this paper proposes a catalog landing zone selection method based on single image.
10929	首先进行障碍检测,并且提取出图像中非障碍区的骨架;	‭First obstacle detection, and extract the skeleton of image china-africa obstacle area.
10930	然后以骨架三线交点为圆心得到多个具有允许着陆范围的着陆圆;	And then get more centered on three skeleton line intersection has allowed scope of landing landing circle.
10931	最后利用计分加权法从着陆圆的半径、连通性、纹理,以及着陆圆心与障碍物的集群关系四个方面得到最佳着陆圆。	Finally using the method of weighted scoring from the landing circle radius, connectivity, texture, and the relationship between landing circle and obstacles of cluster in four aspects: optimal landing circle.
10932	通过与其他方法的对比实验发现,本文方法能选择出障碍物稀少的区域,能够避免着陆器落入地势封闭的环境。	Compared with other methods of experiment found that the method can choose a few obstacles area, to avoid the lander into relief a closed environment.
10933	在由一个源节点、中继节点、目的节点和窃听节点组成的两跳中继窃听信道模型中,假定各节点均具有能量收集能力,采用"储能—发送"模式工作,即先收集能量,然后再发送数据。	By a source node and relay node and destination node and eavesdropping nodes with two-hop relay eavesdropping channel model, the assumption that each node has energy collecting ability, adopt the "energy storage - send" mode, namely first collect energy, and then send the data.