动态Doppler频移下的X射线脉冲轮廓高精度重构方法研究

The X-ray pulsar-based navigation (XPNAV) is a promising yet challenging technique in the field of autonomous navigation, which can provide the spacecraft with the information of position, velocity, time, and attitude. In view of the spacecraft is in dynamic situation, since the detected observation signal has low signal-to-noise ratio (SNR) and its frequency is also unknown, the extraction of pulse profile which is an important observation faces a lot of difficulties and challenges. In order to solve this problem, this proposal is going to develop a pulse profile extraction approach that incorporates the advantages of both the wavelet transform and compressive sensing theories, and to focus on the following key issues: 1) the approach of the high-precision frequency estimation for the non-stationary photon arrival time sequence; 2) the pulse profile reconstruction approach based on the compressive sensing method under low SNR. The main contributions of this proposal are: 1) to establish the mathematical model of the statistical time bin of the photon arrival times and the frequency estimation precision, revealing the coupling relationship between them; 2) to put forward a novel method based on the norm constraint of weighted double-parameters under low signal-to-noise ratio, balancing the big and small coefficients to effectively suppress noise.

X射线脉冲星导航（XPNAV）是一种新型的自主导航方式，可以在单一仪器上自主确定航天器的位置、速度、时间和姿态信息。针对航天器在轨动态情形下，探测器接收的X射线脉冲星信号信噪比低且存在动态Doppler频移，使得脉冲轮廓提取困难的问题。本课题拟研究非平稳光子达到时间序列的高精度频率估计方法，以解决航天器在轨动态Doppler频移的高精度估计问题；研究基于压缩感知的低信噪比下脉冲轮廓重构方法，以实现在信号重构中对噪声分量的有效抑制。两种方法结合共同解决在轨动态下的X射线脉冲轮廓高精度重构的难题，为X射线脉冲星导航技术的应用提供理论基础和方法支撑。本课题的创新性在于：1）提出一种光子到达时间的统计尺度与频率估计精度关系的建模方法，揭示统计尺度与频率估计精度之间的耦合关系，为高精度频率估计方法设计提供参考。2）提出一种基于双参数加权l1范数约束方法，实现在低信噪比下脉冲轮廓的高精度重构。

针对在轨X射线脉冲星观测信号重构和到达时间估计的问题，研究了脉冲星信号辐射特性，明确脉冲星低流量和轨道动态对自转频率估计的影响关系，提出了非平稳随机时间序列的参数估计方法；研究高精度脉冲到达时间估计理论和实验验证方法，构建基于自适应相位谱的动态条件下脉冲到达时间估计模型；开展了在轨X射线脉冲星地面高精度模拟研究，搭建了脉冲星导航地面实验系统，通过实验验证了提出的方法和理论模型。本课题的研究实现了极低流量下非线性区的脉冲星信号估计，突破极低信噪比下时延估计的难题，为深空脉冲星导航和X射线探测器设计提供坚实的基础理论支撑。