X-ray pulsar-based navigation (XNAV) is an ideal navigation mode for deep spacecraft. In order to overcome the time-varying frequency problem caused by the acceleration of spacecraft in the process of photons TOA data, this paper proposes an improved method called Non-uniform epoch folding (NUEF). Specifically, the paper analyzes the mechanism of the traditional EF method, and then divides the traditional method into two key stages: Density Analysis and Digital Filtering. Draws the conclusion that the key to limit the pulse time delay estimation accuracy of the traditional method is that the mathematical expectation of the photon density sequence obtained in the first stage of EF cannot maintain periodicity in a long observation time interval; Then the solution NUEF is given. NUEF method abandons the characteristic that bins uniformly distribute on time axis in traditional EF methods. This method uses the principle of equal phase interval to determine the position and length of each bin by introducing external velocity information, and theoretically overcomes the problem of time-varying frequency in the long observation interval; The simulation experiment verifies the effectiveness of NUEF method. Specifically, the pulse time delay estimation accuracy of NUEF method is improved compared to traditional EF method. And because of the elimination of the periodic search process, the computational complexity of the NUEF method has also decreased.
Aims. The accuracy of the pulsar period estimation directly affects the restoration effect of the signal profile. A more accurate pulsar profile will help improve the accuracy of pulsar delay estimation and thereby improve the performance of X-ray pulsar navigation. This paper proposes a pulsar period estimation method based on photon energy distribution folding and image template matching (PETM). Methods. This method uses the probability distribution information of photon energy for weighted epoch folding. The one-dimensional (1D) profile information was converted into two-dimensional (2D) image information through reverse space-filling curve (SFC) encoding. Then, a feature matching was performed between the target structure and the template structure. At the same time, the criterion of Pearson correlation coefficient (PCC) was used to quantitatively evaluate the matching effect to estimate the optimal period. Results. The simulation results show that the period estimation accuracy of the PETM method is significantly improved, as compared with the traditional χ2-test method. This work also analyzes the folding effect based on the photon energy distribution model and conducts simulation experiments and comparisons on influencing factors, such as noise interference and data quality. At the same time, we also specifically demonstrated the effectiveness of the PETM method for the glitch phenomenon (i.e., a sudden change in period) of pulsar periods. Finally, we also used China’s XPNAV-1 satellite to conduct experiments and analysis of the actual observation data of PSR B0531+21 pulsar within a fixed period of time. The results show that the period estimation accuracy of this method is 4.8190 ns, which is 50.23% higher than the traditional χ2-test method. The method proposed in this article has the advantages of high estimation accuracy, stable estimation performance, strong anti-interference ability, and excellent dynamic period estimation performance. Therefore, it can further improve the navigation performance of X-ray pulsars.
In order to apply the Z2 statistic to the estimation of the navigation pulsar period and time delay, and thus further improve the accuracy of X-ray pulsar navigation, this paper proposes a method for estimating the pulsar period and time delay with complex profiles based on the improved -test. Based on the prior information of the amount of photon data in the observation task, this method adaptively determines the optimal harmonic truncation order under different data qualities. In order to solve the problem of the change rate of complex profile signals being too fast, the binned data is rephased according to the principle of equal photon intensity segmentation within the bin, and the optimal number of bins suitable for the PSR B0531+21 pulsar is selected. Through simulation, a quantitative analysis was conducted on factors that affect the performance of the estimation algorithm, such as observation time, detector area, noise interference, etc. Simulation results show that the proposed estimation method has greater advantages when the observation time is short, the detector area is small, and the interference noise is large. In addition, the observation data of the PSR B0531+21 pulsar is processed and analyzed. The period estimation accuracy of the method proposed is 3.6532 ns, which is 39.57% higher than that of the χ2-test method. The method we proposed has the advantages of being suitable for navigation pulsars, strong environmental adaptability, high estimation accuracy, and strong estimation stability, which can further improve the performance of X-ray pulsar navigation.