on-orbit validation of GPS IIR autonomous navigation
on-orbit validation of GPS IIR autonomous navigation
复制标题
作者：
John A.Rajan
全球定位系统（GPS）是一个国家公用事业，在全球范围内提供导航和定时服务。Block IIR GPS系统由21颗卫星组成，计划在Block II/IIA GPS卫星因故障或性能下降而停止服务时对其进行补充。目前有8颗GPS IIR卫星在轨运行。这些卫星具有使用交叉测距测量的自主自我导航的新能力。这种能力被称为自主导航（AutoNav）。AutoNav设计包括无干扰的后台操作能力。这些在轨运行数据在地面进行后处理，以进行性能验证。本文介绍了在轨数据的性能分析。对在轨数据的分析表明，测量的交联距离存在粗差和精差。粗略的误差是由于世界各地的无意干扰造成的。幸运的是，大多数粗略的错误都可以被识别出来，并在进一步处理中编辑出来。精细误差可能是由于不能正确估计等离子体层传播延迟。补偿等离子体层延迟的原始操作概念是使用双频技术。通过这种技术，AutoNav卫星之间的交叉链路范围在两个不同的频率（fa和fb）下测量。通过比较fa和fb处的交联范围，可以估计和消除由等离子体延迟引起的测距误差。然而，对在轨测试数据的分析表明，测距信号正在经历其他影响，例如由于卫星上的结构而引起的多径。这些其它效应导致测距残差中的精细周期性误差。本文讨论了三种不同的技术，以消除距离残差中的周期性误差的发展和性能。实施了校正算法，校正后的测量值质量足以用于AutoNav处理。校正后的测量值通过AutoNav算法进一步处理，并发现产生了出色的结果。文中还讨论了这三种算法的性能和优缺点。
Global Positioning System (GPS) is a national utility providing navigation and timing service worldwide. The Block IIR GPS system is the set of 21 satellites that are slated to replenish Block II/IIA GPS satellites as they go out of service due to failure or degradation in performance. Currently there are eight GPS IIR satellites on-orbit. These satellites have the new capability to autonomously self navigate using crosslink ranging measurements. This capability is referred to as Autonomous Navigation (AutoNav). AutoNav design includes non-interfering background operation capability. This on-orbit operational data is post processed on the ground for performance validation. This paper presents the performance analysis of the on-orbit data. The analysis of the on-orbit data indicates that the measured crosslink ranges have coarse errors and fine errors. The coarse errors are due to unintentional interference over various areas of the world. Fortunately, most coarse errors can be identified and edited out of further processing. The fine errors are potentially due to the inability to correctly estimate plasmaspheric propagation delays. The original operational concept to compensate for plasmaspheric delays is to use a dual-frequency technique. With this technique, crosslink ranges between AutoNav satellites are measured at two different frequencies (fa and fb). By comparing crosslink ranges at fa and fb, ranging errors caused by plasma delays can be estimated and removed. However, analysis of the on-orbit test data has indicated the ranging signals are experiencing other effects such as multipath due to structures on the satellite. These other effects result in fine periodic errors in the ranging residuals. This paper discusses the development and performance of three different techniques to remove the periodic errors in the range residual. The correction algorithms were implemented and the corrected measurements were of sufficient quality to be used for AutoNav processing. The corrected measurements were further processed by the AutoNav algorithms and were found to yield excellent results. The performance and relative merits of the three algorithms are also discussed.