一种小功率组合式航空发电机

先进的飞行控制系统和机载电子系统设备对轻型飞机电源系统的要求越来越高，电源系统的可靠性将直接影响到轻型飞机的安全性。本文提出了一种适用于轻型飞机的小功率组合式航空发电机方案。这种新型的“永磁─—内封闭导磁体”组合发电机既有永磁发电机的体积小、重量轻、无电刷等优点，又有内封闭导磁体发电机电压可调的优点。在发电机体积重量增加不多的情况下，免除了复杂的直流变换器，使电源系统体积最小、重量最轻，并且提高了系统的可靠性。同时采用晶体管脉宽调制激磁控制方式，使电源系统的稳压性得到提高。     

边缘计算在电机远程运维系统中的应用探讨

介绍了边缘计算的基本概念及与云计算的对比。针对电机远程运维需求,介绍了电机远程运维系统的架构及主要功能。在描述边缘侧设备与远程运维平台的数据交互过程的基础上,详细阐述了边缘计算在电机远程运维系统中的应用。     

CFD软件自动化验证确认云平台设计与实现

针对大型计算流体力学(CFD)软件的验证与确认,为了减少人工成本,提高软件质量和开发效率,并适应于未来高性能计算发展的要求,基于云计算思想,提出了自动化测试云平台的解决方案。该方案采用模块化的浏览器/服务器(B/S)网络架构;利用LAMP(Linux+Apache+MySQL+PHP/Python)开发工具;建立了持续集成的专业数据库;构建了涵盖可靠集群监控、复杂作业调度及大规模并行计算功能的云端环境;实现了通过便捷的网络访问,自动加载测试算例、提交集群计算、监测实时进度、自动化后处理、输出分析结果,并提供丰富的实验对比图、误差分析报表和汇总报告等;完成了对大型CFD软件的自动化验证与确认。将该方案应用于某大规模并行计算的CFD软件,验证了该解决方案的可行性与实用性。     

基于点云数据的发动机管路最小间距计算方案

为检测整机发动机管路是否满足最小间距的设计要求,提出了一种基于点云数据的发动机管路最小间距计算方案,方案包含5个步骤:①从点云数据中划分出不同管路的数据;②基于管路点云数据的空间分布构造等间隔栅格,计算栅格中心点作为管路的趋势线数据;③在管路各趋势线数据点位置上构造垂直平面,将管路点云数据投影到最近的垂直平面上,获得各个垂直平面上呈圆弧状分布的投影点数据;④对各垂直平面上的投影点数据进行最小二乘圆拟合,得到拟合圆圆心及其半径值,将拟合圆圆心作为管路中心线数据;⑤采用遍历法计算两条管路中心线数据的最小间距,中心线最小间距分别减去两条管路的半径值则得到两条管路的表面最小间距。通过12条管路验证了方案的准确度。实验结果表明:管路最小间距偏差在-0.35~0.46mm之间,管路半径偏差在-0.08~0.22mm之间。该方案的实施有助于管路间距数字化检测的实现,且方案的计算结果具有较好的鲁棒性。      

复杂曲面零件散乱点云特征点提取

复杂曲面零件的几何特征提取对加工质量的检测及逆向重构具有重要意义。提出了一种复杂曲面零件散乱点云特征点提取方法。首先,提出了基于高斯权重的邻域主成分分析的方法,通过估计每一点邻域的局部变化程度对点云模型进行初始标记;然后,采用基于标记的自动识别法实现特征点的提取;最后,通过特征点聚类的方式去除了特征点集中的异常点,完善了提取的点云特征。该方法直接操作于散乱点云,无需任何的拓扑连接信息。试验结果表明:该方法简单、有效,不需要过多地人为调节参数,特征点提取完整。     

Satellite and ground detection of very dense smoke clouds produced on the islands of the Paraná river delta that affected a large region in Central Argentina

Intense fires were produced on the Paraná river delta islands, Argentina, during most part of 2008, by a combination of an exceptionally dry period and the farmers’ use of a fire land-cleaning technique. In April 2008, those fires significantly affected the nearby regions and their inhabitants, from Rosario city to Buenos Aires mega-city. In this work we present satellite as well as ground Aerosol Optical Depth (AOD) at 550 nm data obtained during the propagation of pollution clouds to the central zone of Argentina. The highest value (1.18) was registered at Buenos Aires by atmospheric remote sensing, using the satellite instrument MODIS/Terra on April 18th 2008 at 10:35 local time (= UT − 3 h). On the same day, ground air quality detectors also measured in this city the highest Total Suspended Particle (TSP) value of the month, 2.02 mg/m³. The AOD(550) daily variation at Rosario Astronomical Observatory, which is located near the Paraná riverside, was derived by combining solar ultraviolet erythemal irradiance data (measured with a YES biometre) with model calculations. On April 25th 2008, from 12:00 to 15:30 local time, a rather high and constant AOD(550) value was registered, with a mean value of (0.90 ± 0.21). Cities located on the side of the Rosario–Buenos Aires highway (San Nicolás, Baradero and San Pedro) were also affected, showing a mean AOD(550) between the Rosario and Buenos Aires values. The particulate matter was collected with gridded samplers placed on the Paraná river islands as well as at the Rosario Observatory. They were analysed with a Scanning Electron Microscope (SEM) and mainly showed a biological origin. Even if normally large particles travel small distances from the source, organic aerosol in the range of 40–100 μm and complex asymmetric structures were registered several kilometres away from the aerosol sources on the islands. Another event of intense UV index attenuation (98.6%) occurred on September 18th 2008, due to very dense smoke clouds that extended over the Rosario area for several hours. The clouds were driven away from the fires by East–northeast and East–southeast winds. The minimum value of this index measured around noon allows to derive a maximum AOD(550)_max = (3.65 ± 0.90) at 12:45 local time. Soot clouds extended over the Paraná river, transporting Burned Biomass Debris (BBD) that deposited on Rosario. In particular, burned leaves and small branches with dimensions of 1–20 cm were collected. The mean (BBD) particles deposited on the ground from 7:00 to 19:00 local time were (0.92 ± 0.20) BBD/(m² h).     

Geomagnetic activity associated with magnetic clouds,magnetic cloud-like structures and interplanetary shocks for the period 1995–2003

Using nine years (1995–2003) of solar wind plasma and magnetic field data, solar sunspot number, and geomagnetic activity data, we investigated the geomagnetic activity associated with magnetic clouds (MCs), magnetic cloud-like structures (MCLs), and interplanetary shock waves. Eighty-two MCs and one hundred and twenty-two MCLs were identified by using solar wind and magnetic field data from the WIND mission, and two hundred and sixty-one interplanetary shocks were identified over the period of 1995–2003 in the vicinity of Earth. It is found that MCs are typically more geoeffective than MCLs or interplanetary shocks. The occurrence frequency of MCs is not well correlated with sunspot number. By contrast, both occurrence frequency of MCLs and sudden storm commencements (SSCs) are well correlated with sunspot number.     

低空无人机交通管理概览与建议综述

大量无序飞行的低空无人机运行会对地面设施、公共安全、空中载人飞行器等带来危害。然而，目前民航空中交通管理不能适应未来数以百万架的无人机。为了应对该挑战，世界各国针对低空无人机空中交通管理开发了新框架。低空无人机空中交通管理属于近几年出来的新事物。藉此聚焦于低空无人机交通管理，从4个方面进行概览：低空无人机相关的空中交通基本概念及现状、低空无人机交通管理介绍、低空无人机交通管理的关键技术和低空无人机交通管理相关科学问题，以期望对无人机产业健康发展有一定的建议意义。     

激光室内定位的混合式高效点云地图构建方法CSCD

针对机器人在室内定位中存在的点云地图形式单一、存储空间大等问题,提出了一种包含特征地图、通行地图和精简地图的混合形式地图构建方法。构建特征地图时,利用曲率、法线和局部显著性等要素提取环境中的显著特征点。构建通行地图时,首先,采用区域生长分割平面;其次,基于室内曼哈顿假设,利用平面空间关系分割出地平面;最后,根据预设高度构建出2D通行地图,并将3D边缘信息融入到通行地图中。在精简地图中,分别采用主方向权重、随机采样和K均值聚类方法对不同类型体素网格内点云进行精简。实验表明,特征地图可为机器人提供丰富的特征信息。通行地图中地面分割的准确度大于95%,可提供准确的先验通行信息。精简地图有效降低了点云地图的冗余度,在精简比例达到95%时,仍可取得0.8mm的平均模型误差,其精简性能优于传统的随机采样和体素格网方法。     

改进PointNetLK的点云智能配准与位姿图优化方法

针对空间在轨服务任务中的非合作目标相对位姿测量问题,提出一种目标可测部位点云的智能配准方法。首先,通过Straight Through滤波算法对半物理仿真平台采集得到的点云进行目标提取,以消除背景数据等杂乱信息;其次,改进PointNetLK神经网络点云配准算法,将提取后的点云数据作为输入,从而获得初步配准结果,解决非合作目标先验信息缺失导致的无法配准问题;最后,建立基于位姿图的优化模型,以降低配准误差,提高配准精度。实验结果表明,与传统迭代最近点(ICP)算法相比,配准综合误差从6.3598降低到1.7291,精度提高约 72.81% 单次耗时从33.16 s降低到4.2 s,效率提升约87.33%,与当前SM ICP等其他算法相比,也具有一定的优势。