基于K-SVD的星载GPS观测数据压缩方法

The compression method of GPS observation data for satellite navigation based on K-SVD

摘要:

空间飞行器主要采用双频GPS精密定轨技术提高轨道预报精度,但需对双频GPS导航系统的观测数据进行预存储,为星载固存的存储资源增加负担。为此,文章提出了针对星载GPS观测数据的压缩方法,该方法引入稀疏表示理论,采用K奇异值分解(K-Singular Value Decomposition,K-SVD)字典学习方法构建出压缩字典,GPS导航系统将采集的观测数据时间序列稀疏分解到K-SVD压缩字典中,达到数据压缩目的。通过地面仿真验证,该方法不仅具有高压缩比、低失真率优点外,还可以有效降低GPS观测数据的噪声水平,从而提升精密定轨精度,并适用于不同轨道高度的低轨飞行器应用。

Abstract:

Spacecraft mainly use dual-frequency GPS precision orbit determination technology to improve orbit prediction accuracy, but it is necessary to pre-store the observation data of the dual-frequency GPS navigation system, which increases the burden of solid state memory system’s storage resources. So, the article proposes a real-time compression method for satellite-borne GPS observation data. This method introduces the theory of compressed sensing and uses the K-Singular Value Decomposition (K-SVD) dictionary learning method to construct a compression dictionary for GPS navigation. The system sparsely decomposes the collected observation data into the K-SVD compression dictionary to achieve the purpose of data compression. Through simulation verification, this method not only has high compression ratio and low distortion rate, but also can effectively reduce the noise level of GPS observation data and improve the precision of precise orbit determination. The method is suitable for GPS observation data compression application of different orbit.