GNSS高阶电离层延迟改正算法和影响研究

发布时间：2018-03-14 09:25

本文选题：GNSS　切入点：系统硬件延迟　出处：《东南大学》2017年硕士论文　论文类型：学位论文

【摘要】：电离层延迟改正是GNSS导航定位技术中需要解决的关键问题之一。在各种空间物理和地球物理机制的综合作用下,电离层结构及物理化学变化规律十分复杂,电离层对无线电波传播的影响表现为高度的非线性和各向异性,因此有必要深入研究电离层对于无线电信号的影响规律,以消除或者降低电离层误差对GNSS导航定位的影响。本文基于IGS(International GNSS Service)分析中心的数据和地面GNSS接收站的 CORS(Continuous Operational Reference Stations)数据,针对高阶电离层延迟改正问题,主要进行了以下的研究工作:(1)分析了 GNSS系统硬件延迟偏差(Difference Code Bias,简记为DCB)的产生原因,并给出了分离卫星DCB与接收机DCB的通用方法,利用不同纬度上IGS站实测数据,分离了时间跨度为60天的卫星和接收机的DCB,并评估了其稳定性,结果发现:卫星DCB和接收机DCB均是不稳定的,在中高纬度区域硬件延迟信息变化较为平稳,在低纬度地区,硬件系统偏差受地磁和太阳活动影响较大,稳定性明显变差,卫星和接收机的硬件时间延迟的日变化量对于GPS信号L1测距的影响分别可达到2.021TECU和3.085TECU。(2)分析了太阳活动和磁暴现象对高阶电离层效应数值大小的影响,通过顾及高阶电离层延迟的PPP实验,定量地分析了高阶电离层延迟对PPP解算坐标产生的改正值大小,实验发现:在WGS84坐标系下,在太阳与地磁场的活跃期高阶电离层延迟在X、Y、Z轴方向产生的改正值分别为1.89mm、3.90mm、8.45mm,在平静期对应X、Y、Z轴方向的产生的改正值分别为1.2mm、2.40mm、3.91mm,表明无论是在太阳活动与地磁场活动的活跃期还是平静期,电离层延迟高阶项对于mm级的PPP定位的影响都是十分显著的。(3)综合多项式模型与三角级数模型的优点,提出了一种改进的多项式模型,命名为POLY-SIN模型。结合江苏CORS数据,对比了 POLY-SIN模型和传统模型(9-DPM)的建模精度,结果发现:POLY-SIN模型和9-DPM模型的检验中误差分别为 1.181TECU、1.291TECU,相对误差分别为 8.29%、9.39%,POLY-SIN 模型比9-DPM模型精度提高了 8.5%。(4)为了进一步提高区域电离层延迟改正的建模精度,利用H-BP算法对提出的POLY-SIN模型进行误差补偿,构造出一种融合模型,命名为BP+POLY-SIN模型。在同样的实例数据和样本选择下,BP+POLY-SIN模型的检验中误差为1.106TECU,相对误差为7.76%,相比9-DPM和POLY-SIN模型,BP+POLY-SIN模型建模精度最高,较9-DPM模型精度提高了 14.3%。
[Abstract]:Ionospheric delay correction is one of the key problems to be solved in GNSS navigation and positioning technology. The influence of ionosphere on radio wave propagation is highly nonlinear and anisotropic, so it is necessary to deeply study the influence of ionosphere on radio signal. In order to eliminate or reduce the influence of ionospheric errors on GNSS navigation and positioning, based on the data of the IGS(International GNSS Service Analysis Center and the CORS(Continuous Operational Reference stations data of the ground GNSS receiving station, this paper aims at correcting the higher-order ionospheric delay. The main work of this paper is as follows: (1) the paper analyzes the reasons of the hardware delay Code biases of GNSS system, and gives a general method of separating satellite DCB from receiver DCB, and makes use of the measured data of IGS station at different latitudes. The DCB of satellite and receiver with a time span of 60 days is separated, and its stability is evaluated. The results show that both satellite DCB and receiver DCB are unstable, and the hardware delay information varies steadily in the middle and high latitudes, and at low latitudes. The deviation of the hardware system is greatly affected by geomagnetic and solar activities, and the stability of the hardware system becomes worse obviously. The influence of the diurnal variation of the hardware time delay of the satellite and the receiver on the GPS signal L1 ranging can reach 2.021 TECU and 3.085 TECU, respectively. The effects of solar activity and magnetic storm on the numerical magnitude of the higher-order ionospheric effect are analyzed. By taking into account the high-order ionospheric delay in the PPP experiment, the magnitude of the correction value of the high-order ionospheric delay to the PPP solution coordinates is quantitatively analyzed. It is found that in the WGS84 coordinate system, The higher order ionospheric delay in the active period of the sun and the geomagnetic field is 1.89 mm / 3.90 mm / 8.45 mm respectively in the direction of the Xnynez axis, and 1.2mm / 2.40 mm / 3.91 mm in the quiescent period respectively, indicating that both solar and geomagnetic fields are active in the solar and geomagnetic fields. The results show that the correction values of the higher order ionospheric delay are 1.89 mm / 3.90 mm / 8.45 mm and 1.2mm / 2.40 mm / 3.91 mm respectively during the quiescent period. The active period or the quiet period, The influence of higher order term of ionospheric delay on PPP location of mm order is very significant. The advantages of polynomial model and trigonometric series model are synthesised. An improved polynomial model named POLY-SIN model is proposed, which is combined with Jiangsu CORS data. The modeling accuracy of POLY-SIN model is compared with that of traditional model. The results show that the errors in the test of the two models are 1.181TECUU 1.291TECU, respectively, and the relative errors are 8.2999.39. The accuracy of the POLY-SIN model is 8.50.14% higher than that of the 9-DPM model.) in order to further improve the modeling accuracy of the regional ionospheric delay correction, The H-BP algorithm is used to compensate the error of the proposed POLY-SIN model, and a fusion model is constructed. Under the same example data and sample selection, the error of BP POLY-SIN model is 1.106 TECU, and the relative error is 7.76. Compared with 9-DPM and POLY-SIN model, BP POLY-SIN model has the highest modeling accuracy, and the accuracy of BP POLY-SIN model is 14.3 times higher than that of 9-DPM model.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P228.4

【相似文献】