北斗三号共视时间传递性能分析

简要介绍了4种北斗双频组合共视(Common-View, CV)模型,设计并实现了RINEX (Receiver Independent Exchange Format)转换CGGTTS (Common GNSS (Global Navigation Satellite System) Generic Time Transfer Standard)的工具软件rnx2cgg,解决了CGGTTS V2E (Version 2 Extended)不支持北斗三号(BDS-3)新信号体制的问题.系统性地分析了北斗授时误差源,包括空间信号测距误差、伪距测量噪声、差分码偏差,并根据误差模型给出了北斗双频共视的权函数.最后,利用BSNC (Beijing Satellite Navigation Center)和MGEX (The Multi-GNSS Experiment and Pilot Project)观测站数据,开展了北斗授时、站间共视时间传递性能分析.试验结果表明:北斗时从地面段到空间段、用户段的全链路双频共视闭合差的标准差(Standard Deviation, STD)优于2 ns.北...     

基于不同轨道类型BDS卫星的载波相位共视时间传递分析

基于北斗卫星导航载波相位共视(BDS CP(CV))时间传递模型,使用IGS多GNSS试验先导项目(MGEX)提供的精密轨道和钟差产品,研究分析GEO,IGSO和MEO 3种北斗在轨卫星进行时间传递的不确定度。实验结果表明:相对于双向卫星比对结果和光纤比对结果,基于IGSO载波相位时间传递结果的标准差较好,相对于双向卫星比对结果,西安—长春观测数据比对结果的标准差中IGSO较好,可以达到0.39 ns;相对于光纤比对结果,西安—临潼观测数据比对结果的标准差中IGSO较好,可以达到0.18 ns;基于GEO和IGSO的载波相位时间传递结果的稳定度比基于MEO的较好。     

基于“北斗”观测数据的C波段双向卫星时间频率传递中的电离层改正

为了提高C波段双向卫星时间频率传递(TWSTFT)精度,用载波相位观测值平滑伪距的方法,计算电离层延迟对C波段双向卫星时间频率传递的影响。在临潼站和喀什站观测北斗G3星并进行双向比对。该实验结果表明,实验期间C波段双向卫星时间频率传递中电离层延迟影响的最大值为0.47 ns。     

基于IGS MGEX产品分析GPS PPP时间传递

基于Bernese 5.2软件,利用精密单点定位(PPP)的方法,分别采用只包含GPS单系统的IGS快速、最终产品和由IGS多GNSS实验(MGEX)先导项目分析中心GFZ和CODE提供的包含4系统的产品,共4种不同的精密卫星轨道和钟差产品,对PTB-NTSC和PTB-ROB两条国际时间比对链路2个月的GPS数据进行了时间传递试验。选取使用IGS快速产品的时间传递结果为参考,以其他产品时间传递结果与参考值的差异来评价不同精密产品尤其是包含4系统的MGEX产品对GPS PPP时间传递的影响。试验结果表明,不同产品时间传递结果差异的RMS在0.1 ns左右,验证了MGEX多系统产品可以用于GPS PPP时间传递。考虑到MGEX产品本身的自洽性,该结果可以为后续基于MGEX产品进行北斗、Galileo PPP时间传递提供参考。     

IGS产品在GPS时间比对中的应用

在利用GPS CV(GPS Common View)技术进行高精度时间比对时，电离层和卫星位置误差对观测到的卫星信号的影响是不容忽视的，需要对它进行精确的估计和改正．讨论IGS精密星历和CODE全球总电子含量图(TECMAPs)在GPS时间传递中的应用．计算结果表明，采用IGS产品可有效提高单站定时和远距离时间传递的精度。     

多站联合星地时间同步及预报性能分析

无线电双向法通过上行与下行观测量比对实现星载钟与地面站时间同步,比对过程中消除或削弱了绝大多数的公共误差项,因此时间比对精度高。卫星不可视时该方法不能实施,此时卫星钟预报精度只能依赖卫星钟自身的物理性能,不可视弧长越长卫星钟预报精度衰减越快。为了削弱因卫星不可视带来的精度损失,多站联合星地时间同步是一种有效的解决方案。给出了多站联合星地时间同步的基本原理、推导了时间比对模型,并利用COMPASS实测数据分析了多站联合时间同步及预报性能,实验结果表明,多站联合观测有效延长了卫星的可视弧长,为提升卫星钟预报精度提供了数据基础。由于观测设备之间的系统性偏差,各站得到的星地钟差结果可能存在跳变,影响卫星钟预报精度。因此系统性偏差成为制约多站联合星地时间同步性能的关键性因素,系统性偏差的精确标定能确保多站联合星地时间同步及预报精度的大幅提升。     

北斗三号多频点双频组合共视时间传递性能分析

针对北斗二号(BeiDou-2) B1I&B2I标准双频无电离层伪距组合解算的共视比对结果中存在明显的噪声进而影响其短期稳定性的现象, 开展了在BeiDou-2和北斗三号(BeiDou-3)现有可用频率信号中选取最优双频组合的研究, 以期改善BeiDou的共视时间传递性能. 基于开发的CGGTTS (Common GNSS (Global Navigation Satellite System) Generic Time Transfer Standard)软件所生成的标准共视文件, 完成了中国科学院国家授时中心(NTSC)与捷克光电研究院(TP)之间GPS、Galileo、BeiDou-2 B1I&B2I及B1I&B3I和BeiDou-3 B1I&B3I及B1C&B2a双频无电离层伪距组合共视时间比对试验, 并用Vondark滤波对各双频组合的共视结果降噪处理, 通过计算滤波残差的RMS值来评估共视时间比对的精度. 结果表明, 利用BeiDou-3 B1C&B2a组合伪距值获得的共视时间比对结果噪声相对较小, 相比BeiDou-2 B1I&B2I、B1I&B3I和BeiDou-3 B1I&B3I组合的RMS (Root Mean Square)值分别提高约46%、52%和37%, 与GPS P1&P2组合的精度相当, 且与Galileo E1&E5a组合相差不大. BeiDou-3 B1C&B2a组合链路的短稳(< 1d)要优于BeiDou-2 B1I&B2I、B1I&B3I和BeiDou-3 B1I&B3I组合, 且与GPS P1&P2、Galileo E1&E5a组合的稳定性相当; 6条共视时间比对链路的中长稳(> 1d)基本一致.     

基于CGGTTS的北斗3号系统时间性能评估

依据国际时间频率咨询委员会(Consultative Committee for Time and Frequency, CCTF) GNSS (Global Navigation Satellite System)时间比对工作组制定的时间传递标准(Common GNSS Generic Time Transfer Standard Version2E, CGGTTS_V2E), 针对GNSS接收机观测到的伪距信号开发了数据处理软件, 用于生成标准格式的CGGTTS文件, 并对其可靠性进行了验证. 结果表明, 与sbf2cggtts软件生成的CGGTTS文件相比, 在同一历元下, 分别利用相同GPS和BeiDou-2卫星观测值计算的星地钟差值基本一致, 互差绝对值不超过0.5ns的差值分别占总数的96%、94%. 以中国标准时间UTC(NTSC) (Coordinated Universal Time (National Time Service Center))为参考, 利用数据处理软件分别对BeiDou-2和BeiDou-3卫星的B1I和B3I双频消电离层组合观测值处理并生成标准格式的CGGTTS文件, 通过分析其星地钟差参数对BeiDou系统时间的性能进行评估. 结果表明, 与BeiDou-2相比, BeiDou-3系统时间的内符合精度提高约28%, 且1 d以上中长期频率稳定度明显优于BeiDou-2.     

NTSC—PTB链路的GPS全视法时间比对数据分析

采用通过中国科学院国家授时中心(NTSC)GPS单、双频这2类观测设备得到的实测数据,进行了NTSC—PTB(physikalisch-technische bundesanstalt(Germary))链路的全视法比对计算和分析。利用这2类设备得到的GPS AV(全视)的结果与BIPM公布的AV方法的A类不确定度一致,从而确定国家授时中心的2类GPS观测设备的性能达到了国际同类水平。如果对双频观测设备时延校准后,采用P3类型进行比对,有望将NTSC现有的比对精度提高50%。     

TAI 计算的变化

从1999年7月的资料开始,TAI 的计算中,对于长距离 GPS 共视比对结果将不再用电离层时延实测值,改用由国际 GPS 服务(IGS)提供的电离层模型。自 BIPM 采用 GPS 共视规范后,全球参加 TAI 计算的时间实验室的 GPS 共视观测资料按地区分别与 NIST、OP 和 CRL 进行比对(不作电离层改正)。该三站之间的     

电离层TEC对小耀斑的响应

利用国际GPS观测网(IGS)提供的多个台站的观测数据,分析了M级别以下的小、暗太阳耀斑对向阳面电离层TEC的影响．利用传统分析方法的结果表明,从单条视线(LOS)观测数据得到的电离层TEC及其时间变化率曲线来看,由于它们的波动水平和正常情况下的背景电离层变化相当,使此类小耀斑的信息完全淹没在背景噪声中,不能够显示和分辨出耀斑的发生．利用相干求和的数据处理方法,选用向阳面18个GPS台站的观测数据研究了一次C级SF耀斑引起的电离层TEC增加,结果发现,这种方法能有效地消除背景电离层变化噪声,电离层对耀斑的响应非常清楚和明显,这通常只能在X级别的大耀斑中看到．和GOES卫星X射线数据相比,电离层TEC变化的时间特征和耀斑爆发的开始、最大和结束时间均有很好的符合,其最大平均TEC增量在0．1TECU以下,和X级别的大耀斑相比有一个或多个量级上的差别．     

The Performance Evaluation of BeiDou-3 System Time Based on CGGTTS

According to the Common GNSS Generic Time Transfer Standard Version2E (CGGTTS_V2E) developed by the GNSS (Global Navigation Satellite System) Working Group of the International Consultative Committee for Time and Frequency (CCTF), the data processing software is developed by using the pseudorange signal measured by the GNSS receiver, which is used to generate the CGGTTS files in the standard format, and its reliability is verified. The results show that, compared with the CGGTTS files generated by sbf2cggtts software, the offset between the GNSS system time and local time scale calculated by the same GPS and BDS satellite observations in the same epoch is identical, and the difference with the absolute value of the difference less than 0.5 ns accounts for 96% and 94% of the total, respectively. Taking Chinese standard time UTC(NTSC) (Coordinated Universal Time (National Time Service Center)) as the reference time scale, the data processing software is used to process the observations of the B1I and B3I dual-frequency ionospheric combination of BeiDou-2 and BeiDou-3 satellites, and generate the CGGTTS files in the standard format, and the performance of BeiDou system time is evaluated by analyzing the parameter of offset between the GNSS system time and local time scale. The results show that, compared with BeiDou-2, the internal precision of BeiDou-3 system time is increased by about 28%, and the frequency stability of medium and long-term is obviously better than BeiDou-2 after one day.     

Long-term Clock Bias Prediction Based on An ARMA Model

The long-term and reliable prediction of satellite clock bias (SCB) is an important prerequisite for realizing the satellite autonomous navigation and orbit determination. Considering the shortcomings of the quadratic polynomial model (PM) and gray system model (GM) in the long-term prediction of SCB, a new prediction method of SCB based on an ARMA (Auto-Regressive Moving Average) model is proposed to represent the variation characteristics of SCB more accurately. In this paper, a careful precision analysis of the 90-day SCB prediction is made to verify the feasibility and validity of this proposed method by using the IGS (International GNSS Service) clock data. According to the variation characteristics of each satellite clock, the pattern recognition, modeling and prediction of SCB are conducted, and the detailed comparison is made with the other three models at the same time. The results show that adopting the ARMA model can effectively improve the accuracy of long-term SCB prediction.     

通过卫星双向双频观测对电离层时延的测定

介绍了利用卫星双向双频(C波段)观测来测定电离层时延的方法,并对不同经纬度的观测结果进行了比较和分析。卫星双向双频(C波段)观测精度高,采样间隔短,能测定电离层总电子含量的细微变化。     

基于转发式测轨系统的GEO卫星机动检测初探

通过对利用C波段转发式测轨网观测"鑫诺一号"卫星得到的原始资料进行初步分析,探讨卫星轨道机动时卫星在测站与卫星连线方向的距离、速度和加速度的变化规律,并提出一种通过对原始观测数据的拟合比对,实现GEO卫星(地球同步卫星)轨道机动检测的方法。分析比较表明,用该方法检测到的卫星机动开始和结束时刻与星载推力器实际喷火开始和结束时刻之间的差别小于3 min,该方法对处于机动期间的轨道确定有一定的积极意义。     

Electron content near the lunar surface using dual-frequency VLBI tracking data in a single lunar orbiter mission

In VLBI observations of Vstar, a subsatellite of the Japanese lunar mission SELENE, there were opportunities for lunar grazing occultation when Vstar was very close to the limb of the Moon. This kind of chance made it possible to probe the thin plasma layer above the Moon's surface as a meaningful by-product of VLBI,by using the radio occultation method with coherent radio waves from the S/X bands.The dual-frequency measurements were carried out at Earth-based VLBI stations. In the line-of-sight direction between the satellite and the ground-based tracking station where VLBI measurements were made, the effects of the terrestrial ionosphere, interplanetary plasma and the thin lunar ionosphere mixed together in the combined observables of dual-frequency Doppler shift and phase shift. To separate the variation of the ionospheric total electron content(TEC) near the surface of the Moon from the mixed signal, the influences of the terrestrial ionosphere and interplanetary plasma have been removed by using an extrapolation method based on a short-term trend. The lunar TEC is estimated from the dual-frequency observation for Vstar from UT 22:18to UT 22:20 on 2008 June 28 at several tracking stations. The TEC results obtained from VLBI sites are identical, however, they are not as remarkable as the result obtained at the Usuda deep space tracking station.     

Tropospheric Delay from VLBI and GNSS Measurements

Using an updated version of the QUASAR software package developed at the Institute of Applied Astronomy of the Russian Academy of Sciences, we have processed the VLBI observations within the international CONT14 program (May 6–20, 2014), in which a global network of 17 stations was involved (a total of ~250 000 observations). The package update concerned the optimization of data structure and the refinement of stochastic models for the random variations in wet tropospheric delay and atomic clock difference. The main goal of this paper is to compare the VLBI determinations of the tropospheric delay with its independent determinations using global navigation satellite systems (GNSS). We show that both these determinations agree well between themselves only in the case of a global analysis of the VLBI observations, where the VLBI station coordinates are also refined, along with the tropospheric delay and the clock synchronization and Earth orientation parameters. If, alternatively, the station coordinates are insufficiently accurate and are not refined from VLBI observations, then it is appropriate not to determine the tropospheric delay from these observations, but to take it from the publicly accessible independent GNSS data. However, this requires that the VLBI and GNSS techniques operate simultaneously at a common observing site. We have established the shortcomings of the universally accepted method of stabilizing the global solution associated with the absence of a criterion for choosing reference stations and radio sources. Two ways of their elimination are proposed: (i) introducing a coordinated list of weight factors for the errors in the coordinates of such stations and sources into the stabilization algorithm and (ii) adopting a coordinated list of stations and sources the refinement of whose coordinates is not required at all for a certain time.     

A UT1-Like Quantity from Analysis of GPS Orbit Planes

This paper discusses a UT1-like quantity, UTGPS, determined daily from Earth-referenced GPS satellite orbits from the International GPS Service (IGS). For each satellite considered, the observed relation between the satellite's IGS orbit and a model of its orbit plane in inertial space is used to estimate UT1. This modeled orbit plane is initialized using the satellite's IGS orbit and the actual UT1 at an initial time. It is then propagated using standard models of gravitational forces and an empirical model representing the orbit-normal radiation pressure observed during several years of in-flight experience with the satellite. To estimate UT1, an a-priori transformation from terrestrial to true-of-date celestial coordinates is applied to the satellite's IGS orbit. The geocentric angular deviations of points of the resulting transformed orbit from the modeled orbit plane are analyzed, giving the angle between the ascending nodes of the satellite's transformed and modeled orbit planes. To this observed angle between nodes, converted to a UT1 difference, is added the a-priori UT1 value used in the transformation. From the result is subtracted a model of the angle, again converted to a UT1 difference, between the ascending nodes of the actual and modeled orbit planes. The final result is the estimate of UT1 from this satellite, and the median of the UT1 estimates from all satellites considered is UTGPS. The root-mean-square difference between UTGPS-UT1 at the beginning and at the end of an interval of one to four weeks is approximately 30 s times the square root of the interval's duration in weeks.This revised version was published online in October 2005 with corrections to the Cover Date.