| Pulsar is a kind of neutron star with extraordinary high temperature, super-high pressure, ultra-strong magnetic field and excellent stable period. Pulsar is honored as space laboratory with extreme condition, and the most stable clock in the world. In the half century since the discovery of pulsar, the research mainly focuses on astronomy and physics. Recent years, with the development of space technology, people are gradually concerned about the use of pulsar in deep space navigation. XPNAV (X-ray Pulsar-based Navigation) is a new kind of autonomous celestial navigation technology, which can provide position, velocity, time and attitude for the spacecrafts in near-earth, deep space or interplanetary orbit. XPNAV has excellent characteristics, such as strong anti-jamming ability, high reliability, wide applicability, and has significant strategic meanings and huge potential for development.This paper discussed the key problems of XPNAV such as pulsar weak signal processing and simulation experiment system. The main progress of this paper is summarized as follows:1. This paper presents a novel periodicity detection algorithm based on discrete square wave transform (DSWT) and its FPGA-based hardware realization method. Firstly, the periodicity detection using DSWT is proved feasible by the comparison of DSWT and FFT; meanwhile, the white noise suppression of DSWT algorithm is studied. Secondly, because the transform kernel of DSWT only takes +1 or -1, which is more suitable for hardware implementation, this paper presents the algo-rithm's FPGA realization. Lastly, the experiment board is developed using Xilinx Spartan-3 series FPGA chip XC3S2000. The experiment results show that the signal-to-noise ratio of signal which DSWT algorithm can detect is lower than that can be detected by FFT algorithm; the time delay of DSWT algorithm is 3 clocks, which is less than that of FFT algorithm; the hardware resource that DSWT algorithm needed is no more than that of FFT algorithm.2. This paper analyze the characteristic of X-ray pulsar integrated pulse profile, and propose a new recognition algorithm based on selected Bispectra-Mellin (BM) spectra in this paper. Firstly, this algorithm calculates the bispectra of integrated pulse profile to extract shift invariant information and suppress noise. Then, scale stretch of integrated pulse profile is transformed into the phase difference of BM spectra by two-dimensional Mellin transform of normalized amplitude of bispectra. Lastly, simplified Fisher's Separability is introduced to reduce the dimensionality of BM spectra, and the correlation of feature vectors is computed to recognize the type of pulsar. X-ray pulsar data observed by Rossi X-ray Timing Explorer is used in the experiments. The results show that: this proposed algorithm can recognize the integrated pulse profile with different phases and scale factors, and the recognition rate is higher than those of other existing algorithms; the optimal recognition rate is obtained when the phase bin is 4.8 degree; this algorithm has small operation and is suitable for XPNAV system.3. In order to improve the time delay measurement accuracy of integrated pulse profile in X-ray pulsar navigation system, the characteristic of integrated pulse profile is analyzed, and a time delay measurement algorithm based on 1(1/2) spectra is proposed. This algorithm suppresses Gaussian noise using 1(/2) spectra, and transforms the time delay in time domain to phase difference in frequency domain. Minimum mean square error method is introduced to measure the phase difference. Compared with the existing algorithm, this algorithm has the advantage of high measurement accuracy and low computation.4. The time delay measurement algorithm based on 1(1/2) spectra can not suppress the effect of scale stretch, so we propose a new algorithm that combines rough estimation and precise measurement. In the rough estimation, the scale stretch and noise are suppressed by 3rd-order wavelant. Using the parabolic interpolation method of the precise measurement, the iterative operation with small step length is avoided, and the computation of the algorithm is significantly reduced. The experiments are carried out using the pulsar data observed by Rossi X-ray Timing Explorer, and the results show that this algorithm can suppress the scale stretch and noise effectively and the measurement error is smaller than the other existing algorithms; this algorithm achieves the highest measurement accuracy when the phase bin is 0.9 degrees; the computation of this algorithm is less than 10 percent of that of Taylor FFT algorithm, and is suitable for XPNAV.5. A kind of XPNAV simulation experiment system is designed in this paper. This system is composed of two parts: the simulation of pulsar signal and the calculation of navigation parameters. In the first part, the time at which X-ray photon arrive the solar system barycenter is modeled by non-homogeneous Poisson process, then the output pulse of X-ray detector can be simulated by time transformation. In the second part, navigation information included in the pulses which are simulated in the first part can be reached by using Delta-Correction method. This system can simulate five pulsar signals simultaneously, and can be used to study the process of signal processing and parameters calculation of XPNAV in photo level. This system has the advantage of low costs and high simulation accuracy, and provides a reference for the design of prototype in space flight experiment.6. A fault-tolerant integrated navigation system based on XPNAV and strapdown inertial navigation system (SINS) is proposed. The defect of long filter period of XPNAV is remedied by the independency and uninterruptibility of SINS. The long term stability of XPNAV is used to restrain the divergence of SINS errors over time. This system introduces phase residual detection method to detect and isolate the breakdown of XPNAV system which is caused by the shelter from other celestial bodies, and uses federated extended Kalman filter to fuse navigation information. The experimental results show that this method has higher navigation accuracy and shorter filter period than the pulsar navigation method, and is not affected by the shelter from other celestial bodies. |
PDF Full Text Request