Vision-based pose estimation for cooperative space objects

Imaging sensors are widely used in aerospace recently. In this paper, a vision-based approach for estimating the pose of cooperative space objects is proposed. We learn generative model for each space object based on homeomorphic manifold analysis. Conceptual manifold is used to represent pose variation of captured images of the object in visual space, and nonlinear functions mapping between conceptual manifold representation and visual inputs are learned. Given such learned model, we estimate the pose of a new image by minimizing a reconstruction error via a traversal procedure along the conceptual manifold. Experimental results on the simulated image dataset show that our approach is effective for 1D and 2D pose estimation.     

空间目标态势行为与事件本体设计与构建

针对空间目标态势威胁响应和预警研究不完善的问题，为实现空间目标行为和事件的智能推理、实时响应和主动预警，保障航天活动和空间利益，提出面向行为与事件的空间目标态势本体模型(BEO-SO2)。基于基本形式本体(BFO)，构建空间目标态势行为与事件本体模型。在统一时空框架下建立空间目标关系模型，实现空间目标行为执行和事件发生过程的动态推演。以空间目标碰撞威胁为背景，设计对象实体等级体系和碰撞威胁行为与事件要素，建立碰撞威胁推理规则，基于空间目标态势行为与事件本体进行实现、验证与展示。结果表明,空间目标态势行为与事件本体能够实现空间目标碰撞威胁的等级推理与预警，并指导进一步的空间活动，为航天任务与空间活动提供有效保障。     

基于递推差分进化算法的空间机器人参数辨识

针对空间机器人抓取未知目标时的特性参数以及关节摩擦系数的辨识和在线修正问题,提出一种基于递推差分进化算法的实时参数辨识方法。首先采用静态连续摩擦模型描述机器人的关节摩擦特性,并建立两关节空间机器人的非线性动力学模型,然后基于差分进化算法和递推最大似然估计法推导出递推差分进化算法,并用于空间机器人的参数辨识,最后仿真校验了该辨识方法的有效性。仿真结果表明,该算法的辨识精度优于遗传算法和最小二乘法,辨识速度较快,能满足遥操作的要求,且对于动态信息有较好的跟随性。     

Three dimensional spatial memory and learning in real and virtual environments

Human orientation and spatial cognition partlydepends on our ability to remember sets ofvisual landmarks and imagine their relationshipto us from a different viewpoint. We normallymake large body rotations only about a singleaxis which is aligned with gravity. However,astronauts who try to recognize environmentsrotated in 3 dimensions report that theirterrestrial ability to imagine the relativeorientation of remembered landmarks does noteasily generalize. The ability of humansubjects to learn to mentally rotate a simplearray of six objects around them was studied in1-G laboratory experiments. Subjects weretested in a cubic chamber (n = 73) and aequivalent virtual environment (n = 24),analogous to the interior of a space stationnode module. A picture of an object waspresented at the center of each wall. Subjectshad to memorize the spatial relationships amongthe six objects and learn to predict thedirection to a specific object if their bodywere in a specified 3D orientation. Percentcorrect learning curves and response times weremeasured. Most subjects achieved high accuracyfrom a given viewpoint within 20 trials,regardless of roll orientation, and learned asecond view direction with equal or greaterease. Performance of the subject group thatused a head mounted display/head tracker wasqualitatively similar to that of the secondgroup tested in a physical node simulator. Body position with respect to gravity had asignificant but minor effect on performance ofeach group, suggesting that results may alsoapply to weightless situations. A correlationwas found between task performance measures andconventional paper-and-pencil tests of fieldindependence and 2&3 dimensional figurerotation ability.     

基于借力飞行技术的星际多目标交会转移轨道设计

针对深空探测中的转移轨道问题,提出一种星际多目标交会转移轨道设计方法。该方法基于Tisserand原理,采用能量曲线确定交会目标的序列,由Pork-Chop图确定设计参数的可行域和时序,避免了传统方法在交会目标序列和初始轨道段设计中的缺陷。将多目标交会转移轨道设计问题归结为一种非线性多约束多变量搜索寻优问题,通过梯度下降法,选取目标函数的负梯度方向作为每步迭代的搜索方向,逐步逼近目标函数的极小值点,从而使转移轨道优化问题得到简化。最后,将本文的设计方法用于解决欧空局的ROSETTA任务深空转移轨道的设计,设计结果与欧空局公布的结果一致,从而验证了该设计方法的可行性和正确性。     

基于RCS曲线的SAR图像点目标变化检测

地物目标的物理结构、表面粗糙度或地物目标类型发生了变化,则其后向散射能量一 般会发生相应的变化,对应的雷达散射截面(Radar Cross Section,RCS)也会发生变化,这 将导致合成孔径雷达(Synthetic Aperture Radar, SAR)图像的亮度和色泽发生变化。提出 了一种新的基于RCS曲线特性的SAR图像目标变化检测算法。该算法不同于以往的基 于图像域的变化检测算法,从目标的散射特性提取目标的变化信息,避免了不同时相的SAR 图像对误配准所带来的错误。并进行了仿真实验,实验结果表明可行。
     

一种改进的空间载人机动装置参数辨识鲁棒投影算法

空间载人机动装置(MMU)在进行救援过程中,其动力学参数存在很大的不确定性.为解决这一辨识问题,本文首先推导系统关于各动力学参数的线性化模型,然后结合该线性模型的特点,提出了一种用于估计动力学参数的改进鲁棒投影算法,并在理论上分析了该算法的收敛性质,数字仿真验证了方法的有效性.     

Determination of predicted position of a satellite at limited errors of measurements

Non-statistical formulation of the problem of determining the future position of an object in circumterrestrial space orbiting around the Earth is considered at some constraints on measurement errors. Criteria of optimality and the algorithms satisfying these criteria are presented. Computational schemes are given for the most interesting (for practical purposes) central and projective algorithms. Using mathematical simulation methods the precision characteristics of these algorithms are determined at different distributions of measurement errors for typical situations, arising when one supports a catalog of space objects. It is shown that the errors of determining the predicted position of a space object, obtained with the non-statistical approach, can be less than the errors of the least squares method.     

Assessment on the feasibility of future shepherding of asteroid resources

Most plausible futures for space exploration and exploitation require a large mass in Earth orbit. Delivering this mass requires overcoming the Earth's natural gravity well, which imposes a distinct obstacle to any future space venture. An alternative solution is to search for more accessible resources elsewhere. In particular, this paper examines the possibility of future utilisation of near Earth asteroid resources. The accessibility of asteroid material can be estimated by analysing the volume of Keplerian orbital element space from which Earth can be reached under a certain energy threshold and then by mapping this analysis onto an existing statistical near Earth objects (NEO) model. Earth is reached through orbital transfers defined by a series of impulsive manoeuvres and computed using the patched-conic approximation. The NEO model allows an estimation of the probability of finding an object that could be transferred with a given Δv budget. For the first time, a resource map provides a realistic assessment of the mass of material resources in near Earth space as a function of energy investment. The results show that there is a considerable mass of resources that can be accessed and exploited at relatively low levels of energy. More importantly, asteroid resources can be accessed with an entire spectrum of levels of energy, unlike other more massive bodies such as the Earth or Moon, which require a minimum energy threshold implicit in their gravity well. With this resource map, the total change of velocity required to capture an asteroid, or transfer its resources to Earth, can be estimated as a function of object size. Thus, realistic examples of asteroid resource utilisation can be provided.     

航天员微重力作业训练系统重力补偿控制策略

针对航天员微重力作业训练系统的重力场补偿控制这一关键技术,进行了理论和实验研究。分析了模拟微重力环境的机理,确定了微重力作业训练系统的总体结构方案,提出了一种基于电流反馈的重力补偿控制及多干扰力补偿控制策略。通过虚拟重力补偿控制实验,验证了在地面环境、动态作业过程中,模拟物体在不同空间重力加速度环境下的运动规律,实现了在重力方向模拟空间环境下物体移动的作业训练效果。研究成果为在地面实现三维作业训练系统的控制奠定了基础。     

基于改进OOPN的系统工作模式分析方法

针对空间有效载荷系统工作模式的复杂性问题,在研究其动态行为和运行特征的基础上给出了一种基于改进的面向对象Petri网(IOOPN,Improved Obiect Oriented Petri Net)的系统工作模式分析方法.该方法集合了Petri网和面向对象的优点;在面向对象Petri网模型中引入了对象间的行为约束关系;并根据约束关系实现了基于最小序列组的系统工作模式分析算法.最后,将该方法应用于某资源卫星有效载荷系统中,实践结果表明该方法满足对有效载荷系统工作模式的分析、仿真和验证的要求,为有效载荷系统的总体设计和集成测试提供了重要的参考.     

Editorial: Cognitive Semantics and Spatio-Temporal Ontologies

How do people describe the location of a target object to another? This task requires a reference object or frame and terms of reference. Traditional linguistic analyses have loosely organized perspectives around people, objects, or environments as reference objects, using reference terms based on a viewpoint or the intrinsic sides of an object, such as left, right, front, and back or based on the environment, such as north, south, east, and west. In actual communication, social, spatial, and cognitive factors may also affect perspective choice. We examine those factors by varying the spatial information (landmarks and cardinal directions), the communication task (relative cognitive burden to speakers and addressees), and the culture of participants (American and Japanese). Speakers used addressees' perspectives more when addressees had the greater cognitive burden. They also used landmarks and cardinal directions when they were available, especially to avoid difficult discriminations like left/right. Some cases appearing to be perspective taking can be interpreted as using a person as a landmark. Finally, terms like near indicating close proximity were preferred to far and to terms requiring projection of directions. Globally, perspective choices of American and Japanese samples were strikingly similar; that is, Japanese did not select addressees' perspectives more than Americans. The traditional linguistic analyses need to be enhanced to account for effects of cognitive, situational, and social factors.     

Change and continuity in US space policy

2010 saw both the unveiling of a new US National Space Policy and the announcement of a fundamentally different strategy for US human spaceflight that would move from the NASA-government-led Apollo-style approach to a greater reliance on the private sector and international cooperation. This viewpoint puts forward arguments on why change in the US approach to human spaceflight is needed, while acknowledging that achieving it in the face of vested interests and threats to jobs and livelihoods is extremely difficult. It suggests that greater US recognition of the need to ensure the sustainability of space activity (by addressing debris, radio-frequency interference and potential deliberate disruption of spacecraft), and an apparent willingness to countenance international norms to govern space activities, could be the new policy’s most lasting heritage.     

Large-size space debris flyby in low earth orbits

the analysis of NORAD catalogue of space objects executed with respect to the overall sizes of upper-stages and last stages of carrier rockets allows the classification of 5 groups of large-size space debris (LSSD). These groups are defined according to the proximity of orbital inclinations of the involved objects. The orbits within a group have various values of deviations in the Right Ascension of the Ascending Node (RAAN). It is proposed to use the RAANs deviations' evolution portrait to clarify the orbital planes’ relative spatial distribution in a group so that the RAAN deviations should be calculated with respect to the concrete precessing orbital plane of the concrete object. In case of the first three groups (inclinations i = 71°, i = 74°, i = 81°) the straight lines of the RAAN relative deviations almost do not intersect each other. So the simple, successive flyby of group’s elements is effective, but the significant value of total ΔV is required to form drift orbits. In case of the fifth group (Sun-synchronous orbits) these straight lines chaotically intersect each other for many times due to the noticeable differences in values of semi-major axes and orbital inclinations. The intersections’ existence makes it possible to create such a flyby sequence for LSSD group when the orbit of one LSSD object simultaneously serves as the drift orbit to attain another LSSD object. This flyby scheme requiring less ΔV was called “diagonal.” The RAANs deviations’ evolution portrait built for the fourth group (to be studied in the paper) contains both types of lines, so the simultaneous combination of diagonal and successive flyby schemes is possible. The value of total ΔV and temporal costs were calculated to cover all the elements of the 4th group. The article is also enriched by the results obtained for the flyby problem solution in case of all the five mentioned LSSD groups. The general recommendations are given concerned with the required reserve of total ΔV and with amount of detachable de-orbiting units onboard the maneuvering platform and onboard the refueling vehicle.     

Design and development of correlation techniques to maintain a space surveillance system catalogue

A growing interest exists in a future, autonomous European Space Surveillance System (ESSS). Currently, most of the knowledge about Earth-orbiting space objects is based on information provided by the USASPACECOM. This paper presents the required initial orbit determination (IOD) and correlation techniques to process optical measurements. Former studies were focused on the handling of radar measurements, which are summarised with the aim of describing a global procedure for processing hybrid measurement types (combination of radar and optic data for catalogue maintenance). The introduction of manoeuvres are presented due to their importance in the space object catalogue maintenance.The detection of uncatalogued objects and the successful correlation of already catalogued objects involve two different tasks for telescopes: survey and tasking. Assumptions for both strategies are developed on the basis of the previous work developed at the University of Berne (see [T. Flohrer, T. Schildknecht, R. Musci, E. Stöveken, Performance estimation for GEO space surveillance, Advances in Space Research 35 (2005). [1]; T. Flohrer, T. Schildknecht, R. Musci, Proposed strategies for optical observations in a future European Space Surveillance Network, presented in the 36th COSPAR Scientific Assembly (2006). [2]; R. Musci, T. Schildknecht, M. Ploner, Orbit improvement for GEO objects using follow-up observations, Advances in Space Research 34 (2004). [3]; R. Musci, T. Schildknecht, M. Ploner, G. Beutler, Orbit improvement for GTO objects using follow-up observations, Advances in Space Research 35 (2005). [4]; R. Musci, T. Schildknecht, T. Flohrer, G. Beutler, Concept for a catalogue of space debris in GEO, Proceedings of the Fourth European Conference on Space Debris, (ESA SP-587, 2005). [5]]). When a new object appears in the field of view, initial orbit determination must be performed. When only one telescope per site is available, the initial measurements are separated by only a few seconds. Therefore, the initial orbit determination is quite inaccurate due to bad mathematical conditioning of the problem. In order to improve the accuracy of the initial orbit determination, several follow-up observations of the object are required. This implies that the telescope needs to track the detected objects over a long period, and therefore the time available for surveying is reduced. By processing the additional follow-up measurements, separated now by a few hours, the initial orbit determination gives more accurate results and the object can be recovered after an orbital revolution. When several telescopes per site are available, the optical strategies may be modified. The survey tasks can be distributed between the available telescopes. In this way the number of images corresponding to each object increases and to track the detected object over long periods is not always needed. Numerical results will be shown in order to evaluate the accuracy and features of the different telescope strategies. A key point for performing efficiently the cataloguing process is the calculation of the estimated state vector covariance matrix. The covariance matrix analysis allows an adaptive tasking-survey telescope scheduling. Moreover, the implementation of a proper batch orbit determination process by means of a square root information filter (SRIF) requires a realistic initial covariance matrix.Hybrid measurements are available from objects that can be observed through both radar and optical sensors (e.g. GTO objects). The batch orbit determination and correlation process of hybrid measurements is also based on SRIF using an extended measurement model. Both the initial orbit determination methods using radar and optical measurements have to be sufficiently accurate to initialise SRIF correctly. In order to avoid filter divergence, the estimated covariance must be correctly updated after processing both kinds of measurements. The implemented algorithms are explained and their performance is shown through realistic simulations.Techniques to detect and characterise object manoeuvres during the cataloguing process have been developed and implemented. Four main groups of manoeuvre objects have been established by means of their observed permitted orbital ranges (GEO, LEO, MEO–GPS, Molniya). The study is based on the historical TLEs files. When an object with an uncatalogued orbit appears, a comparison between the new orbit and the orbits contained in the permitted ranges of one of the manoeuvre groups is performed. If the required Δa and/or Δi to convert the lost orbit into the detected orbit seems to be feasible, a manoeuvre will be identified and the orbit will be updated in the catalogue. Otherwise, it will be decided that a new object was found. For this purpose, a procedure to estimate the manoeuvres and reset orbits have been developed.     

一种自旋稳定卫星姿态传感器数据异常的诊断方法

针对自旋稳定卫星遥测数据异常时,故障类型和故障源难于辨识的问题,提出一种数据驱动的故障诊断新思路.该方法提出将卫星姿态与传感器之间固有冗余关系是否保持作为判断故障类型的依据,并利用主成分分析把这种关系通过测量空间的特征值进行量化,监测特征值之间相应比例的变化实施判定;借助正交子空间的结构分析,选择SPE统计量进行故障检测和故障源的定位.在遥测数据的基础上模拟风云某型气象卫星的不同故障模式,并应用本文方法进行分析,结果表明,该方法克服了传统贡献图方法无法判别故障类型的缺点,可以实现对故障的正确诊断.     

Space System for Detecting Hazardous Celestial Bodies Approaching Earth from the Daytime Sky (<Emphasis Type="Italic">SODA</Emphasis>)

The concept of the System for the Observation of Daytime Asteroids (SODA system) has been developed, the purpose of which is to detect at least 95% of hazardous celestial bodies larger than 10 m in size that fly towards Earth from the Sun side. Spacecraft, equipped with the optimum version, which has three wide-angle optical telescopes of small aperture (20–30 cm) will be placed in a halo orbit around the L₁ libration point of the Sun–Earth system. This will provide a warning on the hazardous object, approaching from the Sun side, and will allow one to determine the orbit and the point of body entering Earth atmosphere to a sufficient accuracy, at least a few hours before the body collides with Earth. The requirements to the system are considered, the results of a preliminary design of the set of instruments have been described, the areas of visibility are calculated, and the versions of data transmission modes have been proposed. It has been shown that, in cooperation with other (particularly ground-based) projects aimed to observing objects flying from the night sky side, it is possible to detect in advance all hazardous bodies in the near-Earth space larger than 10 m in size that approach Earth from almost any direction.     

空间旋转目标涡流消旋概念与仿真分析

针对消除空间旋转目标姿态旋转的问题,本文将第二代高温超导技术与涡流制动技术相结合,提出超导式涡流消旋的概念。然后,对涡流消旋的技术可行性进行了定性分析,通过对多种消旋技术手段的权衡比较归纳出涡流消旋的优势。接着,采用电磁张量理论,建立了电磁-涡流作用机理的精确磁场和涡流力矩模型;最后,考虑实际消旋的任务需求,对典型的高速旋转目标、低速旋转目标、以及复合旋转目标的消旋进行了动力学仿真研究。仿真结果表明,涡流消旋具有足够强大的制动能力,能有效消除目标的高速旋转或复合旋转运动。     

Mental representation of spatial cues in microgravity: Writing and drawing tests

Humans have mental representation of their environment based on sensory information and experience. A series of experiments has been designed to allow the identification of disturbances in the mental representation of three-dimensional space during space flight as a consequence of the absence of the gravitational frame of reference. This NASA/ESA-funded research effort includes motor tests complemented by psychophysics measurements, designed to distinguish the effects of cognitive versus perceptual-motor changes due to microgravity exposure. Preliminary results have been obtained during the microgravity phase of parabolic flight. These results indicate that the vertical height of handwritten characters and drawn objects is reduced in microgravity compared to normal gravity, suggesting that the mental representation of the height of objects and the environment change during short-term microgravity. Identifying lasting abnormalities in the mental representation of spatial cues will establish the scientific and technical foundation for development of preflight and in-flight training and rehabilitative schemes, enhancing astronaut performance of perceptual-motor tasks, for example, interaction with robotic systems during exploration-class missions.     

火星探测出舱机构的识别定位与坡度测量

为实现火星探测两器分离过程中出舱机构的坡度测量,提出一种基于立体视觉技术的出舱机构识别定位与坡度测量方法。首先,采用特征匹配的方法对场景中的出舱机构进行识别定位。针对巡视器导航相机成像视角变化以及出舱机构倾角变化导致图像产生几何形变,火星表面光照因素的影响以及巡视器硬件的计算能力,并结合出舱机构自身特点,提出一种区域特征与Blob特征相融合的识别定位方法。采用加速鲁棒性特征（SURF）算法检测图像中的局部不变特征,并用最大稳定极值区域（MSER）算法检测图像最稳定极值区域,两种特征相融合之后构造SURF描述算子并进行特征匹配。同时,采用M估计抽样一致性（MSAC）算法计算对应点变换矩阵,实现出舱机构在场景中的初步定位。然后,通过对出舱机构上人工标志的精确定位以及三维重建,并采用主成分分析法进行平面拟合,实现出舱机构的坡度测量。试验结果表明,该方法对成像视角变化以及光照变化具有鲁棒性,提取的有效特征数以及计算时间均优于尺度不变特征变换（SIFT）算法,基本满足火星探测中出舱机构坡度测量要求。