卫星钟性能分析及钟差预报

发布时间：2018-04-04 15:46

本文选题：卫星钟性能　切入点：钟差预报　出处：《西安电子科技大学》2013年硕士论文

【摘要】：卫星钟是卫星导航系统的核心，卫星钟的性能及钟差预报是影响导航定位精度的关键因素，自主导航系统及精密单点定位系统主要靠高精度的卫星钟差预报进行导航定位，卫星钟的性能分析是高精度钟差预报的基础。本文系统地研究了卫星钟性能的指标表征体系，详细分析时域稳定度的各类方差，并明确了其适用范围。从地面模拟和在轨运行两方面分析了卫星钟性能，首先，针对影响原子钟输出频率的各类因素，结合经验公式和地面环境实测数据，通过最小二乘方法确定了时间、光强、温度、铷信号、磁场强度等对频率的影响系数，并发现光强与漂移、钟寿命有直接关系；其次，对卫星钟差在轨数据分析，剔除其整体趋势后，发现卫星钟差数据具有明显的周期性。研究二次多项式、kalman滤波及灰色模型三种算法的预测效果，，分析比较其优缺点，结果表明：kalman滤波具有较好的短期预测效果，二次多项式适合中长期预测，而灰色模型则更适合于长期预测。针对卫星钟噪声特性，改进了kalman的噪声方差矩阵。此外，根据钟差数据的周期性，提出了二次多项式加周期项的预报模型，利用谱分析方法确定该模型中的周期项，并从理论上推导周期项对二次多项式预报算法的改进性能。
[Abstract]:Satellite clock is the core of satellite navigation system. The performance of satellite clock and the prediction of clock difference are the key factors that affect the accuracy of navigation and positioning. Autonomous navigation system and precision single point positioning system mainly rely on high-precision prediction of satellite clock difference for navigation and positioning.The performance analysis of satellite clock is the basis of high precision clock difference prediction.In this paper, the index system of satellite clock performance is systematically studied, the variances of time-domain stability are analyzed in detail, and the applicable range is defined.The performance of satellite clock is analyzed from two aspects: ground simulation and on-orbit operation. Firstly, the time and intensity of the satellite clock are determined by the least square method according to the various factors affecting the output frequency of the atomic clock, combined with the empirical formula and the measured data of the ground environment.The effects of temperature, rubidium signal and magnetic field intensity on the frequency are found to be directly related to the drift and clock lifetime. Secondly, the satellite clock difference in orbit data is analyzed and its overall trend is eliminated.It is found that satellite clock difference data have obvious periodicity.This paper studies the prediction effect of quadratic polynomial kalman filter and grey model algorithm, and analyzes and compares its advantages and disadvantages. The results show that the two-order polynomial filter has better short-term prediction effect, and quadratic polynomial is suitable for medium and long term prediction.Grey model is more suitable for long-term prediction.Aiming at the noise characteristics of satellite clock, the noise variance matrix of kalman is improved.In addition, according to the periodicity of clock difference data, a prediction model of quadratic polynomial plus periodic term is proposed. The periodic term in the model is determined by spectral analysis method, and the improved performance of periodic term on quadratic polynomial prediction algorithm is derived theoretically.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P228.1

【参考文献】