Pseudo-automatic characterization of the morphological and kinematical properties of coronal mass ejections using a texture-based technique

The white light coronagraphs onboard SOHO (LASCO-C2 and -C3) and most recently STEREO (SECCHI -COR1 and -COR2) have detected a myriad of coronal mass ejections (CME). They are a key component of space weather and under certain conditions they can become geo-effective, hence the importance of their kinematic characterization to help predict their effects. However, there is still a lot of debate on how to define the event boundaries for space weather purposes, which in turn makes it difficult to agree on their kinematic properties. That lack of agreement is reflected in both the manual and automated CME catalogs in existence. To contribute to a more objective definition and characterization of white-light coronagraph events, Goussies et al. (2010) introduced recently the concept of “texture of the event”. Based on that property, they developed a supervised segmentation algorithm to allow the automatic tracking of dynamic events observed in the coronagraphs field of view, which is called CORonal SEgmentation Technique (CORSET). In this work, we have enhanced the capabilities of the algorithm by adding several new functionalities, namely the automatic computation of different morphological and kinematic parameters. We tested its performance on 57 well-studied limb CME events observed with the LASCO coronagraphs between 1997 and 2001, and compared the parameters obtained with those from three existent CME lists: two of them obtained from an observer-based detection and tracking method (i.e., two manual catalogs), and the other one based on the automated detection and characterization of the CME events (i.e., a fully automated catalog). We found that 51 events could be tracked and quantified in agreement with the CME definition. In general terms, the position angle, and the radial and expansion speeds are in agreement with the manual catalogs used for comparison. On the other hand, some discrepancies between CORSET and the automated catalog were found, which can be explained by the different delimitation of the CME angular extent.     

无人机航程与续航时间的实时估算方法研究

无人机的航程和续航时间是确定无人机顺利飞行任务的关键战术技术指标。针对无人机任务飞行时,影响其航程和续航时间的因素较多,根据空气动力学原理和运动学原理,探索实时估算无人机航程和续航时间的方法及应用设想。     

线加速度台中稳速台转动对随动台定位稳定性的影响

针对线加速度台用于加速度计测试时,随动台定位受稳速台旋转干扰的现象进行了分析,提出被测设备质量要在负载允许范围内,使设备质心尽可能靠近随动台回转中心,并对其配平,以使稳速台转动带来的干扰力矩大大降低,保证测试中随动台的定位准确度及稳定性。     

差速电机驱动控制系统的研制

本文提出了一种适用于差速电机的驱动控制系统设计方案,该系统由直流转换电路、电机驱动电路和转速监测电路组成。通过样机生产及试验验证,结果表明该系统具有体积小、重量轻、效率高、调速范围宽等优点,满足设计指标和用户使用要求。     

涡破裂诱导的垂尾抖振气动弹性分析

通过试验方法分析了三角翼前缘分离涡与垂尾抖振之间的关系,深入研究了尾迹流动对垂尾抖振各阶模态的激励作用。计算得到了垂尾模型固有频率及各阶模态。在风洞试验中,应用激光片光烟流场显示技术,得到了三角翼模型在风速为30 m/s下,各迎角的涡结构;使用加速度传感器测量了垂尾翼根和翼梢的抖振响应;使用热线风速仪测量了垂尾翼根和翼梢位置的脉动速度分量。结果表明:前缘涡破裂后产生的高湍流度的尾迹是垂尾抖振的直接原因,抖振边界与涡破裂的强度和位置有关;涡破裂后尾迹与垂尾产生共振,使得抖振加速度响应频率与垂尾固有频率一致;涡破裂后,在较小迎角下,尾迹对垂尾的高频振动模态的激励较为明显,在较大迎角下,涡破裂流动对垂尾低频振动模态的激励加强了。     

直升机旋翼转速指示异常故障简析

旋翼转速指示异常是直升机较为常见的故障,本文针对两起较为典型的旋翼转速指示异常故障,从系统原理、故障现象、排除方法等多个方面进行分析并列出故障树,为后续装备保障提供参考。     

面向集群环境的虚拟化GPU计算平台

针对集群系统的多节点多GPU环境,提出一种新型虚拟化GPU计算平台。该平台实现对集群系统所有节点上GPU资源的统一抽象与管理,构建公共GPU资源池。原有GPU应用程序可以不经任何修改而迁移到虚拟化GPU计算平台,并具备访问资源池内任何GPU的能力,编程人员无需显式针对多节点多GPU应用展开MPI编程。应用程序摆脱了单个节点上GPU资源的限制,并具备无差别地访问集群系统中任何可用GPU资源的能力,能有效提高系统总体资源利用率以及吞吐量。采用流水化通信技术,实现对虚拟化GPU计算平台的运行时开销以及节点间数据传输延迟的隐藏。实验表明:与非流水化通信相比,系统总体数据传输延迟降低了50%~70%,具备与节点机本地数据传输等同的通信性能。      

临近空间飞艇艇库外约束及稳定性分析

临近空间飞艇在发放前通常停放在艇库内,一旦离开艇库就不可避免地受到艇库外天气,尤其地面风的影响,不仅影响飞艇发放操作的快捷和安全,而且还事关飞行试验的成败。文章建立了大型飞艇在地面受两侧约束时的力学模型,采用动力学方法,分析了受到地面风作用时飞艇的姿态和约束拉力的变化,得到地面风起时和风速稳定后的解析解,评估了飞艇受到地面风作用时的稳定性和安全性。通过分析得到在两侧约束条件下飞艇对正向风和侧风的承受极限,为临近空间飞艇采用两侧约束方式的可行性提供了判据,也为今后发展临近空间飞艇发放技术提供借鉴。     

基模高斯激光束高倍率扩束实验研究

基于激光光学设计了一个适合实验室环境的高倍率激光扩束系统,并进行了相应的实验研究。实验结果表明束腰半径为0.295mm的激光束,最大被扩束至束腰半径约12.04mm。     

多斜孔壁与机加环气膜冷却燃烧室面换热特性数值研究

数值研究了在燃烧室实际工作条件下,多斜孔壁冷却燃烧室和机加环气膜冷却燃烧室的流动、换热特性,比较分析了内、外壁温度场以及冷却流量的变化.研究发现,采用多斜孔壁后,壁面冷却效果得到强化,相比机加环气膜冷却燃烧室,燃烧室前段壁温略有升高,而后段受到气膜冷却效果增强的影响,壁面温度有所降低,壁面温度梯度明显减小.流量系数明显减小,冷却空气量降低23%,用于组织燃烧用气量增加.多斜孔壁冷却技术可以为高温升燃烧室的发展提供有效的冷却技术.