北斗卫星钟差的CEEMDAN分解与周期项提取方法  

CEEMDAN decomposition and periodic term extraction method of the BDS clock bias

作　　者：梁益丰 许江宁[1] 吴苗[1] 何泓洋[1] LIANG Yifeng;XU Jiangning;WU Miao;HE Hongyang(College of Electrical Engineering,Naval University of Engineering,Wuhan 430033,China)

机构地区：[1]海军工程大学电气工程学院,武汉430033

出　　处：《中国惯性技术学报》2022年第4期476-484,共9页Journal of Chinese Inertial Technology

基　　金：国家自然科学基金(41804076)。

摘　　要：卫星钟差具有非线性非平稳特征,其趋势分量、周期分量等信号与随机分量互相干扰,不利于钟差估计与预报、时间尺度算法等实际应用。为了实现卫星钟差数据的信噪分离与重构,提出了一种以自适应噪声完备集合经验模态分解(CEEMDAN)为基础的信噪分离方法。经CEEMDAN分解得到多个本征模态分量,引入排列熵算法确定噪声主导分量与信号主导分量,通过t检验分析分量的高频与低频特征,从而完成信号重构。对各类北斗卫星原子钟进行实验,结果表明,以CEEMDAN为基础的综合分析方法能够准确辨识本征模态分量特征,重构周期项的频谱图中基本不含有不规则周期分量与多余信号;去除周期分量后,4颗北斗卫星原子钟的万秒频率稳定度平均提升21.6%。Satellite clock bias has nonlinear and nonstationary characteristics,its trend component and periodic component interfere with each other,which make against to the practical application of clock difference estimation and prediction and time scale algorithm.In order to realize the signal and noise separation and reconstruction of satellite clock bias data,the signal and noise separation method based on the complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)is proposed.Multiple eigenmode components are obtained by CEEMDAN decomposition,the permutation entropy algorithm is introduced to determine the noise dominant component and the signal dominant component,and the high frequency or low frequency characteristics of the components are analyzed by t-test to complete the signal reconstruction.For all kinds of BDS atomic clock,experimental results show that the comprehensive analysis method based on CEEMDAN can accurately identify the intrinsic mode component characteristics,the spectrum map of the reconstructed periodic items basically does not contain irregular periodic components and redundant signals.After removing the periodic components,the frequency stability of the four Beidou satellite atomic clocks increases by 21.6%on average.

关 键 词：北斗卫星导航系统 卫星原子钟 自适应噪声完备集合经验模态分解 频率稳定度 

分 类 号：TN976[电子电信—信号与信息处理]