基于自回归滑动平均模型的大口径天线风速预测方法

针对大口径、高性能射电望远镜天线受到的随机及时变风扰的问题, 利用自回归滑动平均模型预测望远镜周围风速, 提前计算风致结构变形量, 同时为望远镜伺服控制系统提供足够执行时间来降低风扰影响. 基于新疆奇台110m口径全向可动射电望远镜(QiTai Telescope, QTT)台址风场数据特征, 通过赤池信息准则和贝叶斯信息准则辨识模型阶次, 利用最大似然法估计模型参数, 分析模型残差特性以检验自回归滑动平均模型的有效性. 通过计算不同季度预测数据与测试数据偏差得到预测模型的精度, 夏季平均绝对误差为0.133mcdots^-1, 秋季平均绝对误差为0.162mcdots^-1, 冬季平均绝对误差为0.287mcdots^-1. 整体来看, 基于QTT台址不同季度风速数据建立的自回归滑动平均模型预测误差较小, 满足射电望远镜风扰控制系统的需求, 能为大口径射电望远镜风扰控制提供必要数据支撑.     

一种基于元学习的大口径射电望远镜俯仰轴承故障辨识方法研究

在经过长期运行后大口径射电望远镜俯仰轴会出现微小扭曲, 滚动轴承作为承载俯仰轴的核心部件, 也会因长期承受交变载荷增加疲劳风险, 导致轴承寿命以及望远镜指向精度的下降, 极大影响望远镜的性能. 以俯仰轴承为研究对象, 开展故障辨识方法研究, 可为望远镜天线的高性能运行提供重要支撑. 为实现在有限数据和复杂工作条件下准确地辨识俯仰轴承故障, 提出了一种小样本条件下基于元学习的故障辨识方法(Few-shot Meta-learning Fault Identification, FMFI). 首先将不同工况下的原始信号转换为时频图像数据, 之后按照元学习协议将数据样本随机采样到不同的学习任务中. 在有限样本的条件下, FMFI可以通过训练任务中的样本信息获取通用的先验知识, 在未知的测试任务下实现准确快速的故障辨识. 选取了与望远镜俯仰轴承工况具有相似性的变负载轴承数据集进行实验, 实验结果表明, FMFI方法具有很高的准确性和可靠性, 为大口径射电望远镜俯仰轴承的主动运维和高质量服役提供了有力的技术支持.     

基于数值模拟的射电望远镜台址测风塔位置优选方法研究

射电望远镜天线结构设计和观测任务的正常进行都需要台址风场信息.随着望远镜性能要求的提高,天线结构设计越来越复杂,风载荷对天线观测时造成的指向影响也越发明显,如何在设计时保证结构刚度和强度,在观测时提高有效观测时间,都需要准确的风场数据.由于传统台址测风塔布置方法无法对测风塔拟设点的可靠性做出定量化评估,因此提出一种基于数值模拟优化测风塔位置的方法.数值模拟基于规范参数设置边界条件,与实测数据比较,整体趋势吻合,满足精度需求.对试验台址设置了4个测风塔位置,经分析P2点与天线位置的风速均方根误差值(root-mean-square error, RMSE)最小,测得的数据更能表征天线区域的风场特性.     

基于扇形子镜的12m光学红外望远镜拼接共相误差仿真分析

为满足国内对应用大型通用光学望远镜在光学红外波段进行天文观测的需求, 中国计划建设一架12 m大口径光学红外望远镜(12-meter large aperture optical-infrared telescope, LOT). LOT对于中国天文界具有极其重要的意义, 将会在近红外波段实现共相. LOT的主镜将采用子镜拼接的方案, 由72块扇形子镜拼接而成, 并且中央部分有一块直径4.5m的圆形镜. 分析了采用扇形子镜拼接方案下LOT的共相误差, 提供了详细的共相误差理论推导过程, 并对包括3种子镜面外误差在内的误差源进行深入的单独模拟和综合模拟. 在共相波长为1μm的条件下, 当服从正态分布的倾斜误差的均方根值小于0.016^\prime\prime或服从正态分布的活塞误差的均方根值小于42.5nm时, LOT的斯特列尔比大于0.8. 仿真结果对于采用共相拼接镜面主动光学的主镜在LOT中的应用具有重要意义.     

环焦天线口径面相位分析

环焦天线具有特殊的电磁特性和应用领域.对环焦天线的口径面相位误差进行了理论和仿真分析,推导了馈源和副面位置偏差引起的相位误差、主副面之间的补偿关系以及全息测量中天线转动引起的光程差.研究结果将对环焦天线的精确面形测量和补偿提供理论依据和参考.     

基于斩波轮技术的K波段接收机噪声校准研究

接收机是射电天文中用于探测微弱射电信号的重要接收设备.接收机的强度校准就是将接收机对射电源的响应转换为天文意义上的流量密度.常规方法就是使用经典的冷热负载法,将接收机自身的强度响应转换为一个等效的温度值,之后再据此对射电源做进一步标定.通过搭建基于斩波轮技术的K波段接收机强度校准平台,使用斩波轮法测试K波段常温接收机的噪声温度,并与传统冷热负载法的测试结果进行比对.结果显示,在晴好天气条件下,斩波轮法在30°、90°仰角下噪声温度的最大测试误差为7.5%和8.4%,可以很好地应用于实际噪声温度测试中;但在5°仰角测试中,由于过低仰角引入了地面噪声,使得斩波轮法的测试误差上升至20%–30%之间而无法使用.希望在此基础上进一步开展K波段天空亮温度的理论计算与实测,从而完善斩波轮技术的应用,使之可以满足在不同气象条件下的噪声校准测试需求.     

基于SE-Inception-v3的星系形态分类模型

随着天文探测技术的快速发展, 海量的星系图像数据不断产生, 能够及时高效地对星系图像进行形态分类对研究星系的形成与演化至关重要. 针对传统的星系形态分类模型特征选择困难、分类速度慢、准确率受限等难题, 提出一种以Inception-v3神经网络为主干结构, 融合压缩激励(Squeeze and Excitation Network, SE)通道注意力机制的星系形态分类模型. 该模型在斯隆数字巡天(Sloan Digital Sky Survey, SDSS)样本的测试集准确率高达99.37%. 旋涡星系、圆形星系、中间星系、雪茄状星系与侧向星系的F1值分别为99.33%、99.58%、99.33%、99.41%与99.16%. 该模型与Inception-v3、MobileNet (Mobile Neural Network)和ResNet (Residual Neural Network)网络模型相比, SE-Inception-v3宽度和深度优势表现出更强的特征提取能力, 可以高效识别不同形态的星系, 为未来大型巡天计划的大规模星系形态分类问题提供了一种新方法.     

图像移位堆叠方法用于暗天然卫星的天体测量

暗弱天然卫星与主带小行星相比,具有亮度低、速度变化快的特点.在观测这类天体时,不能简单地延长曝光时间来提高其信噪比.尝试观测多幅短曝光的CCD (chargecoupled device)图像,采用移位堆叠(shift-and-add)方法,希望提高目标成像的信噪比,获得暗弱天然卫星的精确测量结果.使用2018年4月9—12日夜间,中国科学院云南天文台1 m望远镜(1 m望远镜)拍摄的木星5颗暗卫星的229幅CCD图像,实施了移位堆叠试验.为了验证结果的正确性,与相近日期中国科学院云南天文台2.4 m望远镜(2.4 m望远镜)观测的相同木卫图像的测量结果进行了比较和分析.位置归算采用了JPL (Jet Propulsion Laboratory)历表.结果表明,对CCD图像使用移位堆叠方法,通过叠加约10幅曝光时间100 s的图像, 1 m望远镜能观测暗至19等星的不规则天然卫星,而且测量的准确度与2.4 m望远镜的测量结果有良好的一致性.     

GALILEO搜救定位处理的分布式数据融合方法

对遇险目标搜索救援的支持是GALILEO全球卫星导航系统的一项重要民用服务.文章介绍了GALILEO搜索救援子系统(SAR)定位的基本原理.在GALILEO SAR中测量量为时间信号与频率信号,文章详细地给出了两种测量信号的定位方法.基于多传感器信息融合技术,对搜索救援系统中不同测量信号进行综合状态估计.在仿真计算中,采用带噪声的模拟测量数据进行定位,数值计算的结果显示基于多传感器数据融合方法的综合状态估计是可靠的.     

基于NPSVM模型的超新星识别方法

巡天观测与高能物理、黑洞天文等领域均有密切的联系.基于星系-超新星二分类问题,研究光谱数据预处理,结合余弦相似度改善PCA(Principal Component Analysis)光谱分解特征提取方法,用SDSS(the Sloan Digital Sky Survey)、WISeREP(the Weizmann Interactive Supernova data REPository)组成的5620条光谱数据集训练支持向量机,可以得到0.498%泛化误差的识别模型和新样本分类概率.使用Neyman-Pearson决策方法建立NPSVM(Neyman-Pearson Support Vector Machine)模型可进一步降低超新星的漏判率.     

Analysis and Correction of the Influence of Track Irregularity on the Antenna Pointing

According to the influence mechanism of the antenna track irregularity on the telescope pointing accuracy, the distribution of the track errors and their influence on the pointing of the Urumqi Nanshan 26 m telescope are reanalyzed after the antenna track was reformed by using the whole-body welding technology, and hereby the pointing error model is correspondingly revised. By using the moving least-squares method, the measured height errors of the antenna track plane are fitted with a closed curve, and the tilt of the antenna azimuth axis caused by the track irregularity can be determined accordingly. Comparing it with the measured deviation of the antenna azimuth axis caused by the deformation of antenna pedestal, it can be found that both deviations are strongly correlated. A new pointing error model is established in view of the gravity deformation of antenna pedestal, which includes the north-south and east-west components, as well as the antenna track irregularity. Finally, by scanning a known calibration radio source at different positions in the sky, the measured pointing errors are fitted with the new pointing error model. The result shows that the sinusoidal component of the model error can be well constrained by the new pointing correction model, indicating that the new model can reflect very well the antenna pointing error, and can amend it to a certain extend.     

Architectures for science: Intelligent agent technology

I review the current architecture of the HTN and make three suggestions for the future. (i) We should retain the expertise split between agents which deal with the science programmes and those which deal with telescope constraints. This makes it easy to add new programmes or new telescopes. (ii) We should develop “look ahead” schedulers which attempt to schedule a whole night at once. This will give reliable calculations for the chance an observation will be carried out, and give a better chance that high priority time critical observations are successfully scheduled. (iii)We should strive to attract more science programmes to the HTN, in particular time critical observations spread over many nights, and non‐time critical work which can benefit from access to databases and the literature. (© 2008 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

An RTML transport implementation on the Liverpool Telescope

The Liverpool Telescope is a 2.0m fully robotic telescope based on the Canary island of LaPalma. In this paper I describe the Java(TM) implementation of a method to transport RTML (Remote Telescope Markup Language) between a Node Agent (Web Services endpoint) embedded at the telescope and a remote Agent. The advantages and disadvantages of the method are discussed, along with the suitability of the protocol for HTN standardisation. (© 2008 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Sodium beacon tip–tilt determination with Rayleigh-aided auxiliary telescope technique

The usefulness of tracking the Rayleigh portion of a mesospheric sodium laser guide star as reference for absolute tip–tilt recovery in the frame of the auxiliary telescopes technique is shown. This approach leads to the reduction of the ground occupation needed to attain a given sky coverage by more than one order of magnitude. Speed, tracking precision, and the number of auxiliary telescopes are also reduced, making this new approach a more attractive one. The use of a low-altitude Rayleigh spot reinforces the fundamental limitations affecting this and other techniques, thus degrading significantly the quality of the recovered tip–tilt. However, it is shown that, provided adequate care is taken in the collection and treatment of data, an interesting tilt signal can still be retrieved.     

Grid integration of robotic telescopes

Robotic telescopes and grid technology have made significant progress in recent years. Both innovations offer important advantages over conventional technologies, particularly in combination with one another. Here, we introduce robotic telescopes used by the Astrophysical Institute Potsdam as ideal instruments for building a robotic telescope network. We also discuss the grid architecture and protocols facilitating the network integration that is being developed by the German AstroGrid‐D project. Finally, we present three user interfaces employed for this purpose. (© 2008 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

A Study on the Paul-Baker System as an Instrument for Multi-object Spectrum Observation

Paul-Baker systems with 4°flat field and 5°flat field are studied. Their light obstructions under different f/ratios of the primary mirror are analyzed. Due to the strong f/ratio of the system, a focal length extender is designed in order to match the following fiber instrumentation, and two kinds of dispersion prism correctors are designed for correcting the atmospheric dispersion. We compare the designed Paul-Baker system with LAMOST, the national major scientific project now under construction.