基于改进指针网络的卫星对地观测任务规划方法

针对卫星观测任务规划问题约束复杂、求解空间大和输入任务序列长度不固定的特点,使用深度强化学习（DRL）方法对卫星观测任务规划问题进行求解. 综合考虑时间窗口约束、任务间转移机动时间和卫星电量、存储约束,对卫星观测任务规划问题进行建模. 基于指针网络（PN）的运行机制建立序列决策算法模型,使用Mask向量来考虑卫星观测任务规划问题中的各类约束,并通过Actor Critic强化学习算法对模型进行训练,以获得最大的收益率. 借鉴多头注意力（MHA）机制的思想对PN进行改进,提出多头注意力指针网络（MHA-PN）算法. 根据实验结果可以看出,MHA-PN算法显著提高了模型的训练速度和泛化性能,训练好的MHA-PN算法模型可以直接对输入序列进行端到端的推理,避免传统启发式算法迭代求解的过程,具有较高的求解效率.     

密集观测场景下的敏捷成像卫星任务规划方法

针对密集观测场景下敏捷成像卫星任务规划问题求解空间大、输入任务序列较长的特点,综合考虑时间窗口约束、任务转移时卫星姿态调整时间、存储约束和电量约束,对敏捷成像卫星任务规划问题进行建模. 提出融合IndRNN和Pointer Networks的算法模型（Ind-PN）对敏捷成像卫星任务规划问题进行求解,使用多层的IndRNN结构作为算法模型的解码器. 基于Pointer Networks机制对输入任务序列进行选择,使用Mask向量考虑敏捷成像卫星任务规划问题中的各类约束. 基于Actor Critic强化学习算法对算法模型进行训练,以获得最大的观测收益率. 实验结果表明,对于密集观测场景下的任务规划,Ind-PN算法的收敛速度更快,可以获得更高的观测收益率.     

一种适合卫星可靠组播的差错恢复方案*

针对卫星网络高误码率、长延时等特点,提出了一种适合卫星可靠组播的差错恢复方案。该方案采用异或（XOR）编译码的方法,可以一次重传恢复多个丢失的数据包;在此基础上,设计了一种基于重传矩阵预处理的最优重传策略来实现重传数量最小化。仿真结果证明了该可靠组播差错恢复方案在卫星环境中的有效性。     

Bone Marrow-Mesenchymal Stromal Cell Secretome as Conditioned Medium Relieves Experimental Skeletal Muscle Damage Induced by Ex Vivo Eccentric Contraction

Bone marrow-mesenchymal stem/stromal cells (MSCs) may offer promise for skeletal muscle repair/regeneration. Growing evidence suggests that the mechanisms underpinning the beneficial effects of such cells in muscle tissue reside in their ability to secrete bioactive molecules (secretome) with multiple actions. Hence, we examined the effects of MSC secretome as conditioned medium (MSC-CM) on ex vivo murine extensor digitorum longus muscle injured by forced eccentric contraction (EC). By combining morphological (light and confocal laser scanning microscopies) and electrophysiological analyses we demonstrated the capability of MSC-CM to attenuate EC-induced tissue structural damages and sarcolemnic functional properties’ modifications. MSC-CM was effective in protecting myofibers from apoptosis, as suggested by a reduced expression of pro-apoptotic markers, cytochrome c and activated caspase-3, along with an increase in the expression of pro-survival AKT factor. Notably, MSC-CM also reduced the EC-induced tissue redistribution and extension of telocytes/CD34⁺ stromal cells, distinctive cells proposed to play a “nursing” role for the muscle resident myogenic satellite cells (SCs), regarded as the main players of regeneration. Moreover, it affected SC functionality likely contributing to replenishment of the SC reservoir. This study provides the necessary groundwork for further investigation of the effects of MSC secretome in the setting of skeletal muscle injury and regenerative medicine.     

基于GF_1卫星的丹江口水库水面面积-蓄水量-水位相关性研究

基于GF_1卫星16m分辨率多光谱遥感数据及水文观测数据,通过计算NDWI水体指数来提取丹江口水库水面面积,并研究水面面积、水位与蓄水量三者之间的相关性,实现动态分析水库蓄水量变化情况,以期为"南水北调"提供决策支撑。研究发现:(1)南水北调中线首个调度年前后丹江口水库水面面积-水位,水位-蓄水量以及水面面积-蓄水量之间均为显著相关关系,且三对变量依次呈现明显的1次、2次、3次曲线关系,拟合曲线方程的决定系数分别达到0.976、0.999和0.980,因此可以根据其中一个已知变量预测其他两个未知量;(2)水面面积自2014年10月5日至2016年5月5日最大变化幅度相差180km~2,是影响周边环境变化的重要因素;(3)同时蓄水量在首个调度年内,于2014年11月初达到最大值204.10亿m~3,可调水量在16.53~78.93亿m~3范围内浮动,且年内整体变化平缓稳定,保证了2015调度年内南水北调中线工程的供水需求。     

存在飞轮安装偏差与故障的卫星姿态容错控制

基于滑模控制理论提出一种刚体卫星姿态稳定控制方法,实现对反作用飞轮部分失效故障的容错控制,以及对其安装偏差的补偿控制,并从理论上证明了所设计的控制器在有限时间内能够实现对闭环姿态控制系统的几乎全局渐近稳定控制.最后将所设计的控制器应用于某型卫星姿态稳定任务,仿真结果验证了所提出方法的有效性.     

基于短消息平台的气象卫星业务维护管理系统

为满足静止气象卫星地面应用系统的运行管理需求,保证业务系统的高可靠性和维护响应的高时效性,设计了一套基于短消息平台的业务维护管理系统。该系统由短消息平台、实时业务监测和自动维护管理三部分组成,能够实时监测多套卫星系统的业务关键点,及时发现业务故障和特殊事件,利用短消息平台实现消息的集中管理和即时发送,同时具备故障策略管理、维护调度管理和自动化维护响应等功能。实施效果表明该系统具备良好的适用性和可扩展性,能有效减轻维护管理人员的工作压力,为高时效和高可靠要求的大型系统维护管理工作提供有效的技术手段。     

Depth and temperature selection of lake charr (Salvelinus namaycush) ecotypes in Lake Superior revealed by popup satellite archival tags

Lake charr exhibit morphological diversity in large North American lakes, largely attributed to habitat partitioning. Bathythermal habitats of lean lake charr have been assessed but remain largely unknown for other lake charr ecotypes. Popup satellite archival tags (PSATs) were used to determine depth and temperature profiles of lean (n = 15), siscowet (n = 16) and redfin (n = 3) lake charr ecotypes in Lake Superior during segments of the year. Monthly median depths of leans were < 20 m throughout the year while the median depth of redfins was 24–32 m (June-September). Monthly median depth of siscowets ranged from 103 to 204 m but they exhibited four distinctive patterns: deep (>80 m), shallow (<10 m), extreme vertical movements (>80–0 m), and high frequency vertical movements between 100 and 125 m. Siscowets were the most stenothermal with habitat temperatures generally 4–5 °C but not exceeding 12 °C. Leans were found as low as 0 °C (January-March) and up to 15 °C (July-September). Median monthly habitat temperatures for redfins were 1–3 °C colder than leans during July-September. PSATs confirmed that siscowets exhibit extreme vertical movements and may feed near the surface. The timing and duration of the extreme vertical movements exhibited by siscowets did not appear to be associated with time of day (i.e., not diel vertical movements) and may be described as opportunistic movements presumably related to foraging. Data obtained from the PSATs reinforces the view that lake charr are highly adaptive and can thrive in cold, oligotrophic lakes by optimizing the use of the entire water column.     

基于资源三号卫星数据的数字正射影像图制作研究

资源三号卫星的成功发射填补了我国立体卫星测绘的空白,可以提供优于5m地面分辨率的遥感影像.论文提出基于资源三号卫星遥感数据的数字正射影像图（DOM）制作流程,重点对影像纠正、影像配准、影像融合、影像镶嵌等关键步骤进行研究,并通过实验验证方法的可行性.实验证明该方法具有较强的实用价值和广泛的应用前景.     

一种评估LEO卫星通信网络抗毁性的新方法研究

抗毁性是低轨卫星通信网络的重要属性,评估卫星网络抗毁性对卫星网络结构设计具有重要意义.针对低轨卫星网络的特点,在对复杂网络抗毁性的研究基础上,建立以自然连通度为基础的卫星网络抗毁性测度指标,利用STK/Matlab仿真工具分析卫星网络的抗毁性.实验分析表明,仿真结果可以有效评估卫星网络抗毁性大小.