舵面电动加载系统的自适应CMAC复合控制

针对无人机舵面电动加载系统具有非线性及多余力矩的特点,提出了一种自适应CMAC(Cerebellar Model Articulation Controller)神经网络与自适应神经元控制器并联构成复合控制结构.该控制策略以系统的指令输入和实际输出作为CMAC的激励信号,以系统的当前控制误差作为CMAC的训练信号.提出了利用误差在线自适应调整学习率的方法,消除了常规前馈型CMAC的过学习和不稳定现象.建立了无人机舵面电动加载系统的数学模型,给出了具体的控制结构和算法.仿真结果表明:该方法有效抑制了加载系统的多余力矩,增强了系统的稳定性,明显改善了舵面电动加载系统的动态性能.     

无人机舵面负载模拟系统的小脑模型控制

为解决无人机舵面负载模拟系统中非线性和多余力矩扰动问题,利用小脑模型神经网络非线性逼近能力强、结构简单、适于实时控制等特点,采用小脑模型和传统PD(Proportional-Derivative)控制结合的复合控制策略,由小脑模型实现前馈控制,PD控制实现反馈控制,以保证在系统运行各阶段的控制精度.分析讨论了复合控制的不稳定性问题,研究了基于可信度分配和学习率自适应调整的改进型小脑模型的应用情况,提出一种适用于单输入单输出系统的简化小脑模型复合控制设计方法.仿真结果表明该方法有效地解决了小脑模型和PD复合控制的不稳定问题,改善了系统动态加载性能,并具有很好的抗干扰性能.     

基于加性对偶四元数的惯性/天文组合导航算法

为了提高惯性/天文组合导航系统在高动态条件下的导航精度,提出了一种基于加性对偶四元数的惯性/天文组合导航算法.该算法将载体的旋转和平移统一起来,使用螺旋矢量更新对偶四元数,同时补偿圆锥误差和划船误差.推导了组合导航系统基于加性对偶四元数的误差模型和导航参数误差的计算方程;把陀螺仪和加速度计的常值误差扩充到状态变量中,随机误差作为系统噪声输入,利用星敏感器输出参数来校正陀螺漂移,通过卡尔曼滤波对状态变量进行估计.仿真结果表明:在高动态条件下,基于对偶四元数的惯性/天文导航算法的导航精度比传统算法提高2倍多.     

基于加性对偶四元数的惯性/天文组合导航算法

为了提高惯性/天文组合导航系统在高动态条件下的导航精度,提出了一种基于加性对偶四元数的惯性/天文组合导航算法.该算法将载体的旋转和平移统一起来,使用螺旋矢量更新对偶四元数,同时补偿圆锥误差和划船误差.推导了组合导航系统基于加性对偶四元数的误差模型和导航参数误差的计算方程;把陀螺仪和加速度计的常值误差扩充到状态变量中,随机误差作为系统噪声输入,利用星敏感器输出参数来校正陀螺漂移,通过卡尔曼滤波对状态变量进行估计.仿真结果表明:在高动态条件下,基于对偶四元数的惯性/天文导航算法的导航精度比传统算法提高2倍多.     

可重复使用飞行器进场着陆拉平纵向控制

针对可重复使用飞行器（RLV）进场着陆拉平段的纵向控制问题,提出了反馈控制与前馈控制相结合的复合控制策略。设计反馈控制律参数,在此基础上基于时间加权高度跟踪误差/误差变化率平方积分指标优化设计前馈控制律参数。按输入补偿的前馈控制在不影响系统稳定性的前提下,提高了RLV对拉平着陆轨迹的跟踪精度,减小了RLV的接地散布。提出了基于积分器初值的控制律平滑切换方法,实现了RLV起落架释放前后不同控制律之间的平滑切换。仿真验证了拉平纵向复合控制和拉平过程中不同控制律之间平滑切换方法的有效性。      

Lyapunov自适应滤波及其 在INS/SMNS组合中的应用

为克服在惯导/景像匹配导航系统(INS/SMNS,Inertial Navigation System/Scene Matching Navigation System)组合导航系统中随着匹配点增多最小二乘法易出现发散的缺点,提出了一种基于多输入多输出(MIMO,Multi-Input-Multi-Output)系统的改进型Lyapunov自适应滤波算法.该方法利用观测方程信息,通过定义滤波误差和预测误差获得算法的递推公式,而算法的稳定性和收敛性可由Lyapunov稳定性理论来证明.以INS/SMNS组合导航系统为例进行了仿真验证,和递推最小二乘算法相比该算法简单、可靠,而且滤波的性能优于递推最小二乘算法.     

一种改进的最小二乘景像匹配算法

为提高景像匹配系统的匹配速度和对初始定位误差、噪声的鲁棒性,对实时图与基准图之间的灰度分布关系进行分析,建立实时图和基准图之间的景像匹配模型,给出了一种改进的最小二乘景像匹配算法.该算法在最小二乘目标函数中引入一个辅助约束项构成综合目标函数,辅助约束项隐含有对量测输入平滑性的约束,提高了匹配算法的稳定性,运用牛顿法推导出该算法的递推公式,该算法充分利用了综合目标函数的一阶、二阶导数信息,因此具有较快的收敛速度.仿真结果表明了算法的有效性.     

一种基于DCT的SAR图像自聚焦算法

研究了一种基于离散余弦变换(DCT, Discrete Cosine Transform)的SAR(Synthetic Aperture Radar)图像自聚焦算法.该方法从复图像域出发,通过在距离压缩相位历史域引入相位误差模型,改变图像的聚焦程度直至一维像的DCT序列在高序数区域的权值达到最大,从而完成误差校正.同相位梯度自聚焦算法相比,该方法无需在图像域分离出强点目标,因此特别适用于无任何明显特征的图像.由于DCT存在快速算法,使得该自聚焦算法计算量较少,更易实现.实测数据及仿真数据的成像结果证明了此方法的可行性.      

基于权值分配及多特征表示的在线多示例学习跟踪

针对复杂环境下目标跟踪过程中由于遮挡、目标姿势及光照条件变化引起跟踪漂移的问题,提出一种基于多示例学习（MIL）框架的在线视觉目标跟踪算法。该算法针对多示例跟踪算法采用单一haar-like特征不能准确描述目标外观变化及在学习过程中对样本包中各正负样本示例采用相同权值,忽略不同正负样本示例在学习过程中对包的重要性不同的特点,采用多特征联合表示目标外观构造分类器,通过将多特征互补特性融入在线多示例学习过程中,利用多特征的互补属性建立准确的目标外观模型,克服在线多示例跟踪算法对目标外观变化描述不足的问题;同时,依据不同正负样本示例对样本包的重要程度进行权值分配,提高跟踪精度。实验结果表明,本文跟踪算法对场景光线剧烈变化、遮挡、尺度变化及平面旋转等干扰具有较强的跟踪鲁棒性,通过对不同视频序列进行测试,文中算法在5组测试视频序列上的平均中心位置误差远小于对比增量式学习跟踪,仅为10.14像素,其对比算法IVT、MIL和OAB的中心位置误差分别为17.99、20.29和33.64像素。      

Research on the Relation Between the Seeker's Handover and the Geometrical Position of the Missile and the Target

 研究了雷达导引头的交班误差以及截获概率与交班时刻弹目几何位置的关系,提出了单个误差角的导引头天线指向角误差模型,它与两个误差角的导引头指向角误差模型相比可以更好地分析天线指向角误差,并得到使指向角误差最小的弹目几何位置;然后进一步得到使多普勒频率误差最小的弹目几何位置;给出了最终的截获概率的表达式,截获概率由目标处于导引头天线主波束的概率、目标回波多普勒频率落入接收机频带内的概率和目标的检测概率组成的;最后分析了一种典型的中末交班情况.     

Comparison of the magnetic field before the subsolar magnetopause with the magnetic field in the solar wind before the bow shock

Crossings of the magnetopause near the subsolar point are analyzed using data of THEMIS mission. Variations of the magnetic field near magnetopause measured by one of THEMIS satellites are studied and compared with simultaneous measurements in the solar wind by another THEMIS satellite. The time delay of the solar wind arrival at the subsolar point of the magnetopause is taken into account. 30 and 90 s averaging of the magnetic field in the magnetosheath is produced. The results of averaging are compared with the results of measurements in the solar wind before the bow shock and foreshock. It is shown, that _Bx$B_x$ component of the magnetic field near magnetopause is near to zero, which supports the possibility to consider the magnetopause as the tangential discontinuity. Comparatively good correlation of _By$B_y$ component in the solar wind and near the magnetopause is observed. The correlation of _Bz$B_z$ component near the magnetopause and IMF is practically absent, the sign of the _Bz$B_z$ near the subsolar point does not coincide with the sign of IMF _Bz$B_z$ in ∼$\sim$30% cases.     

Consequences of the application of the streamer fluid model to the study of the sprite inception mechanism

We present the results of a streamer-fluid model used to investigate the electrodynamical coupling between the troposphere and upper atmosphere due to the penetration of lightning electric fields into the mesosphere and the lower ionosphere, generating sprites. The model solves the continuity equation for electrons and ions coupled to Poisson equation. The dominant physical response of the atmosphere is the formation of a screening-ionization wave. The wave shields the atmosphere above it from the action of the lightning field and, together with the conductivity reduction below it due to attachment, the wave amplifies the total field below it, allowing for the penetration of intense electric fields in the mesosphere as it propagates downwards into regions of higher density that compress the wave. This is the key physical mechanism for sprite inception. We evaluated the effects of the thundercloud charge geometry, lightning current waveshape, atmospheric conductivity, via different electron density profiles, and the effect of ionization, attachment and electron mobility coefficients in the electrical breakdown process, related to halo production, and sprite streamer initiation. The results showed that electrons with higher mobility are more efficient in shielding the lightning electric field before breakdown, causing delay, and they contribute to the formation of the streamer seed after breakdown, anticipating the sprite streamer inception. Similarly, a higher effective ionization rate, produced by modifications in the attachment and ionization coefficients, anticipates sprite inception. The simulations with 6 different electron density profiles, and therefore conductivities, spanning 4 orders of magnitude, showed that the altitude of breakdown and sprite initiation, as well as their time delays from the lightning discharge are directly related to atmospheric conductivity: higher conductivities produce halo and sprite inception at lower altitudes with longer delays and may hinder sprite formation. We document that variations of 30 times in the lightning current leads to sprite initiation altitudes in the range 66.0–73.5 km, with delays between 1.550 and 34.500 ms, while variations of 4 orders of magnitude in the conductivity profile lead to initiation altitudes 61.0–70.6 km, with delays in the range 3.825–9.825 ms. Consequently, we suggest that lightning characteristics dominate over atmospheric parameters in determining sprites’ initiation altitude and delay. The simulation of a −CG, with a constant current of 30 kA, did not produce a sprite seed, confirming an asymmetry in the response of the atmosphere to positive and negative lightning. This is due to the free electron drift direction that is away from the screening ionization wave, preventing the formation of the streamer seed for the great majority of −CGs. The same does not apply to halos, which depend on the occurrence of breakdown and can be produced by discharges of both polarities.     

Method for the prediction of the effective dose equivalent to the crew of the International Space Station

This paper describes a methodology for assessing the pre-mission exposure of space crew aboard the International Space Station (ISS) in terms of an effective dose equivalent. In this approach, the PHITS Monte Carlo code was used to assess the particle transport of galactic cosmic radiation (GCR) and trapped radiation for solar maximum and minimum conditions through an aluminum shield thickness. From these predicted spectra, and using fluence-to-dose conversion factors, a scaling ratio of the effective dose equivalent rate to the ICRU ambient dose equivalent rate at a 10 mm depth was determined. Only contributions from secondary neutrons, protons, and alpha particles were considered in this analysis.     

On the correlation of the solar wind observed at the L5 point and at the Earth

The L5 point is a promising location for forecasting co-rotating high-speed streams in the solar wind arriving at the Earth. We correlated the solar wind data obtained by the Nozomi spacecraft in interplanetary space and by the Advanced Composition Explorer (ACE) at the L1 point, and found that the correlation is significantly improved from that of the 27-day recurrence of ACE data. Based on the correlation between the two spacecraft observations, we estimated the correlation of the solar wind velocity between the L5 point and at the Earth, and found that the correlation coefficient was about 0.78 in late 1999, while that of the 27-day recurrence was 0.51. Eighty-eight percent of the velocity difference falls within 100 km/s between the L5 point and the Earth. This demonstrates the potential capability of solar wind monitoring at the L5 point to forecast the geomagnetic disturbances 4.5 days in advance.     

Possible physical properties of the flow in the vicinity of the heliopause

An interface between the fully ionized hydrogen plasma of the solar wind (SW) and the partially ionized hydrogen gas flow of the local interstellar medium (LISM) is formed as a region where there is a strong interaction between these two flows. The interface is bounded by the solar wind termination shock (TS) and the LISM bow shock (BS) and is separated on two regions by the heliopause (HP) separating the solar wind and charged component of the LISM (plasma component below). The BS is formed due to the deceleration of the supersonic LISM flow relative to the solar system. Regions of the interface between the TS and HP and between the HP and BS were in literature named as the inner and outer heliosheaths, respectively. An investigation of the structure and physical properties of the heliosheath is at present especially interested due to the fact that Voyager-1 and Voyager-2 have crossed the TS in December 2004 (Burlaga, L.F., Ness, N.F., Acuna, M.Y., et al. Crossing the termination shock into the the heliosheath. Magnetic fields. Science 309, 2027–2029, 2005; Fisk, L.A. Journey into the unknown beyond. Science 309, 2016–2017, 2005; Decker, R.B., Krimigis, S.M., Roelof, E.C., et al. Voyager 1 in the foreshock, termination shock and heliosheath. Science 309, 2020–2024, 2005; Stone, E.C., Cummings, A.C., McDonald, F.B., et al. Voyager 1 explores the termination shock region and the heliosheath beyond. Science 309, 2017–2020, 2005) and in September 2007 (Jokipii, J.R. A shock for Voyager 2. Nature 454, 38–39, 2008; Gurnett, D.A., Kurth, W.S. Intense plasma waves at and near the solar wind termination shock. Nature 454, 78–80, 2008. doi: 10.1038/nature07023; Wang, L., Lin, R.P., Larson, D.E., Luhmann, J.G. Domination of heliosheath pressure by shock-accelerated pickup ions from observations of neutral atoms. Nature 454, 81–83, 2008. doi: 10.1038/nature07068.14; Burlaga, L.F., Ness, N.F., Acuna, M.H., et al. Magnetic fields at the solar wind termination shock. Nature 454, 75–77, 2008. doi: 10.1038/nature07029; Richardson, J.D., Kasper, J.C., Wang, C., et al. Cool heliosheath plasma and deceleration of the upstream solar wind at the termination shock. Nature 454, 63–66, 2008. doi: 10.1038/nature07024; Stone, E.C., Cummings, A.C., McDonald, F.B., et al. An asymmetric solar wind termination shock. Nature 454, 71–74, 2008. doi: 10.1038/nature07022; Decker, R.B., Krimigis, S.M., Roelof, E.C., et al. Mediation of the solar wind termination shock by non-thermal ions. Nature 454, 67–70, 2008. doi: 10.1038/nature 07030), respectively, and entered to the inner heliosheath.     

太阳风——磁层相互作用的磁流体力学数值模拟研究

磁层位于地球空间的最外层, 太阳风与磁层的相互作用是空间天气变化因果链中承上启下的关键环节, 是揭示地球空间天气基本规律的关键科学课题. 地球空间由于时变、多成分、多自由度的关联相互作用,使得传统的理论分析变得非常困难. 数值模拟作为近几十年发展起来的一个新的研究手段,对地球空间的理论和应用研究产生了深刻的影响. 国际上磁层的全球MHD数值模拟工作开始于20世纪70年代末, 最初的研究局限于二维空间. 由于磁层内在的三维特性, 20世纪80年代三维MHD数值模拟工作兴起. 本文简要说明了三维全球磁层MHD (磁流体力学)研究的特点及现状, 给出了三维全球磁层模型的基本框架, 综述了行星际激波与磁层相互作用、大尺度电流体系、重联电压和越极电位、磁层顶K-H不稳定性等方面的太阳风--磁层相互作用的MHD数值模拟的研究进展.      

Influence of the Ionosphere on the Observation of the Mid-Latitude Pi2 Pulsations at the Global Scale

Results of our investigation showed that occurrence frequency of Pi2 over a 24 hour period undergoes seasonal variations in time coincidence with foF2. In the winter months, at sunrise and sunset (when foF2gradients are the largest) the observation probability of these oscillations is minimal. At periods of summer solstice when the F2-layer persists almost round the clock, no effect of Pi2 pulsation attenuation is observed at sunrise and sunset. The pulsation amplitudes behave in a similar manner. Results of this study suggest the conclusion that the propagation of signal from the Pi2 sourse into the mid-latitudes, and also the parameters of these pulsations are essentially affected by electron density in the ionospheric F2-layer.      

辐射变色薄膜剂量计性能测试及加速器应用

本课题的目的是建立用于辐射加工高剂量水平吸收剂量测量的辐射变色薄膜剂量计。通过系统研究,批量制备了以尼龙为基材,副品红氰化物为染料的辐射变色薄膜剂量计。为了验证批制辐射变色薄膜剂量测量的可靠性,选择了国际上应用较广的FWT-60膜及丙氨酸薄膜剂量计与本实验室批制的辐射变色薄膜开展实验室内剂量比对,比对结果均在±4%内符合,归一化偏差En绝对值均小于1。通过在加速器上进一步应用实验表明批制辐射变色薄膜可用于电子束辐照参数测量。     

落塔微重力水平激光干涉测量方法中的干扰因素分析

落塔是获得微重力环境的重要设施,落塔微重力水平的测量对微重力科学实验的研究至关重要. 激光干涉是测量落塔微重力水平的一种新方法,这种方法的基本原理是让一个参考落体在落舱中自由下落,落舱由于受到服外空气阻力的作用将与自由落体运动略有差异,1 11. 肫内的参考落体则更接近理想的自由落体运动,这就使得落舱与参考落体之间存在着加速度差,这种加速度差便反映了落服的微重力水平,其所导致的相对运动则可通过激光干涉的方法测量出来. 本文对落塔微重力水平的激光干涉测量方法中将会遇到的一些主要干扰因素进行了分析,计算结果表明,这些干扰网素所造成的总误差约为 1.2 x 10-7g,低于微重力水平的预测值 10-4~lO-6 g,因此该方法是一种比较可行的测量方法.