| X-ray pulsar navigation has become one of the most potential astronomical navigation methods because of its unique advantages.X-ray pulsar positioning and velocimetry navigation is a new research direction of X-ray pulsar navigation.First,the research status is studied in China and abroad,then the basic theory of X-ray pulsar positioning and velocimetry are introduced,and then the key technologies such as the simultaneous estimation of X-ray pulsar pulse time of arrival(TOA)and period,SNR estimation of pulsar observation profile and navigation filtering are studied respectively in this dissertation.Finally,there are summaries of the full paper and prospect of the next work.The main contents are as follows:1.In view of the influence of Doppler velocity in pulse TOA estimation and the large amount of calculation in most of the period estimation methods,an ultra-fast simultaneous estimation of pulse TOA and period based on the genetic algorithm optimized quantum compressed sensing(GQCS)are proposed using the formation mechanism of pulsar distortion profile.First,the multi-distortion and multi-phase profile dictionary is constructed in GQCS,which is adopted as the probability amplitude of quantum.The quantum measurement matrix is formed by multiple quantum observations of the profile dictionary.Then,the genetic algorithm with memory mechanism is used to optimize the measurement matrix to obtain an optimized one with smaller size and better performance.Finally,the distortion and phase shift of pulsar observation profile,which are pulsar period and pulse TOA,are obtained by using super-resolution matching estimation.Simulation results show that GQCS has higher estimation accuracy and less computation than classical 2method.And GQCS realizes the simultaneous estimation of pulse TOA and period.2.The signal-to-noise ratio(SNR)of pulsar observation profile determines the design of navigation filter.In order to accurately estimate the SNR of the profile,two estimation methods of the SNR of pulsar observation profile based on random forest are proposed for pulsars with different phase shifts,namely,wavelet aided random forest(WA-RF)and bi-spectrum random forest(BSP-RF).WA-RF uses the wavelet low-frequency coefficients of pulsar profiles with different SNR to train the random forest,and estimate the SNR of the observed profile using the trained random forest.Because of the time shift invariance and strong anti-noise ability of bi-spectrum transform,BSP-RF uses the bispectrum of pulsar profile to replace the wavelet in WA-RF.The simulation results show that the SNR estimation accuracy of WA-RF is higher than that of BSP-RF for high flow pulsars without phase shift;For pulsars with phase shift or low flux,the SNR estimation accuracy of BSP-RF is higher than that of WA-RF.3.In order to improve the navigation accuracy of X-ray pulsar positioning and velocimetry system,the direct/indirect hybrid velocimetry pulsar navigation is proposed by combining the characteristics of high flux and low flux pulsars.The hybrid velocimetry pulsar navigation uses one X-ray detector to observe one high flux pulsar and two low flux pulsars in turn.For the high flux pulsar,the direct velocimetry measurement model is used to calculate the two measurements of positioning and velocimetry directly.For low flux pulsars,the positioning measurements are estimated in the indirect velocimetry measurement model using the cross-correlation method.The measurement deviation model based on Doppler effect is used as the positioning compensation which is the way to obtain the indirect velocimetry measurements.Combined with the orbit dynamics model of spacecraft and the measurement model of three pulsars,the Kalman filter is updated in turn,and finally the three-dimensional positioning and velocimetry of spacecraft are realized.The simulation results show that compared with the indirect velocimetry pulsar navigation,the positioning and velocimetry accuracy of the hybrid velocimetry pulsar navigation are improved by 23.35% and 22.74%,respectively. |