基于容器云的地面测控资源池日志管理系统的研究

新型航天地面测控装备基于商用服务器集群，采用虚拟资源池架构和容器云技术，按需进行测控资源调度和测控业务能力生成。和众多服务设备一样，日志是系统故障诊断的主要途径，由于海量日志可能分散在资源池系统中的不同位置，传统的流式日志采集和分析模式难以大批量管理信息，经常导致系统异常发现不及时、排查效率低、定位难度大等问题，日志的检索和统计缺乏行之有效的手段，弊端明显。针对这种地面测控资源池系统的运维管理需求，本文设计了一种基于容器云的日志管理系统，将平台日志和应用日志聚合管理，根据用户需求和运维目的分类检索并定制可视化日志显示，为使用者精准分析系统异常、快速进行系统维护提供有效支撑。     

超参数自适应的MOEA/D-DE算法在翼型气动隐身优化中的应用

MOEA/D-DE 算法易于实现，被广泛应用于处理多目标优化问题，但其超参数CR 和F 对算法性能影响较大。基于MOEA/D-DE 算法框架、利用Sobol 全局灵敏性分析方法对差分进化算子中的交叉控制参数CR进行改进，使用莱维飞行策略控制比例因子F，使算法中的超参数拥有自适应能力，得到超参数自适应的MOEA/D-DE 算法——MOEA/D-DEAH 算法；对MOEA/D-DEAH 算法、不同超参数设置的MOEA/D-DE算法和NSGAII 算法进行函数测试和翼型气动隐身优化算例对比。结果表明：MOEA/D-DEAH 算法性能良好，具有较强的鲁棒性，气动隐身优化效果也比其他算法更好。     

Coupled evolution of piston asperity and cylinder bore contour of piston/cylinder pair in axial piston pump

The wear condition of the piston/cylinder pair is crucial to the performance and reliability of the axial piston pump. The hard piston surface, the soft cylinder bore surface, and the interface oil film affects each other during the wear process. Specifically, in the mixed lubrication region, the geometry of the hard piston surface asperity directly affects the wear of soft cylinder bore surface, while the asperities may deform or even degrade when penetrating and sliding against the cylinder bore. So far, there is no suitable method to simulate their coupled evolution. This paper proposed a wear process simulation model considering the real-time interaction between the elasto-plastic deformation of the piston surface asperity, the wear contour of the cylinder bore, and the lubrication condition of the interface. An offline library of the elasto-plastic constitutive behavior of the asperity based on the finite element method (FEM) is established as a part of the simulation model to precisely analyze the deformation and degradation of the asperity and quickly invoke them in the numerical wear process simulation. The simulation and experimental results show that the piston asperity and the cylinder bore contour converge to a steady state after running-in for about 0.5 h. The distribution of the simulated asperity degradation and wear depth is also verified by the experiment.     

Quantitative contribution of climate change and anthropological activities to vegetation carbon storage in the Dongting Lake basin in the last two decades

The Dongting Lake Basin is an important hydrological regulation and flood storage area in the Yangtze River Basin, which plays an important role in maintaining regional ecological security. The watershed vegetation and its carbon sequestration capacity have changed dramatically due to climate change and human activities in the last two decades. In this paper, the monthly and annual vegetation net primary productivity (NPP) of the Dongting Lake basin during 2000 to 2020 was firstly estimated using the improved Carnegie-Ames-Stanford Approach (CASA) model. Then the multi-year NPP change trend and its significance were analyzed based on Theil-Sen median and Mann-Kendall method. Subsequently, the Hurst index was used to simulate the vegetation NPP persistence in the study area. Finally, the driving mechanisms of vegetation NPP changes in the study area were explored using partial correlation coefficients and residual analysis. The results demonstrated that: 1) The annual average NPP in the basin showed a fluctuating and increasing trend from 273.54 to 718.30 g C/m²·yr¹ during 2000 to 2020, and except for autumn, all seasons (spring, summer, winter) and annual NPP series showed an upward trend. The spatial distribution of vegetation NPP is characterized by an asymmetrical horseshoe shape in general, with high values in the west, south and east parts, and low values in the Lake area; 2) During the study years, about 84.38 % of the basin area showed a significant and extremely significant increase of NPP; 3) The future trend of vegetation NPP in the basin shows that the area of decrease (22.79 %) is more than the area of increase (11.35 %). The Chang-Zhu-Tan urban agglomeration will generally show a continuous and extremely significant reduction trend, while parts of Dongting Lake will show a continuous and extremely significant increase trend; 4) The correlation between solar radiation and NPP is stronger than the other two meteorological factors (precipitation and temperature). Temperature has a significant inhibitory effect on NPP, solar radiation has a significant promoting effect on NPP, and the effect of precipitation on NPP was relatively complicated. The relative importance of meteorological factors on vegetation NPP was ranked as follows: solar radiation > precipitation > temperature. The impacts of both climate change and human activities on NPP changes showed great spatial variability, and the positive contributions (89.81 % for climate change and 82.87 % for human activities) were both greater than the negative contributions.     

变循环发动机风车起动建模及仿真研究

为了满足变循环发动机风车起动性能仿真的需求，建立了变循环发动机部件级风车起动模型。针对旋转部件等熵效率不连续的问题，提出使用换算扭矩代替等熵效率的方法，给出了旋转部件全转速特性拓展方法。提出了考虑点火及燃烧稳定性的燃烧室稳定性模型。考虑了变循环发动机的8个可调参数，采用差分进化算法对变循环发动机的风车及风车起动性能进行了优化。结果表明，风车状态时，变循环发动机在单外涵模态具有更高的核心机物理转速，有利于点火之后核心机物理转速快速趋于其慢车值。单外涵模态时，变循环发动机在风车状态的可调参数仅与飞行马赫数有关，为了保持较高的核心机物理转速，后涵道引射器外涵面积需随飞行马赫数的增加而减少，其余参数皆固定在其最佳值。飞行高度6km、马赫数0.8时，变循环发动机在单外涵模态下的风车起动时间为1.4s。风车起动过程中燃油流量的增长主要受燃烧稳定性所约束。通过对可调参数的优化，可使变循环发动机在风车起动过程中的关键性能参数最大程度的逼近其限制值，从而减少风车起动时间。     

一种基于实测点云数据的壁板类零件修配加工路径计算方法

针对飞机壁板类零件采用修配法装配过程中需要精确获取修配加工路径的问题，提出一种基于实测点云数据的壁板类零件修配加工路径计算方法。首先设计一种基于法线差的特征点检测算法提取参考壁板的初始特征点。然后建立迭代收缩优化模型对初始特征点进行收缩优化，得到参考壁板的准确特征点。最后根据工装定位孔特征进行配准，将参考壁板的特征点映射到待修配壁板上，并将映射后的特征点按照一定顺序进行连接，得到待修配壁板的修配加工路径，按照修配路径进行加工，获得壁板装配配合边界。实验表明，该方法能够精确提取待修配壁板修配加工路径，并且经过加工验证，壁板试验件对缝间隙获得小于1.2 mm的实际效果。     

基于条纹投射三维测量的铆钉自动化检测技术

针对大型薄壁结构铆接点位自动化检测问题,提出了基于条纹投射三维测量的铆钉检测技术,实现了铆钉镦头尺寸特征高精度测量以及裂纹缺陷自动化识别。在传统条纹投射三维测量的基础上,引入高动态范围(HDR)纹理合成,提出二、三维结合的点云分割策略,实现铆钉尺寸特征高效提取。搭建深度学习神经网络,实现镦头表面缺陷识别。本文搭建了系统原理样机,以铆接桁条样品和标准器作为样件进行系统实验验证。结果表明:实验室条件下,该方法直径测量精度为0.040 mm,高度测量精度为0.013 mm,缺陷识别准确率可达99.30%。与传统方法相比,本文提出的方法更加易于集成,效率高,具有广泛应用价值。