Research On Differential Positioning Based On Regional Hybrid Constellation Navigation Satellite System And Integrated Navigation Among Constellations

Hybrid constellation navigation satellite system which contains several types of orbit has powerful ability to overlay specific region. At present, the navigation algorithms always suit well for Walker constellation navigation system, in which the satellites share many characters, such as the same orbit height, the same tiny eccentricity. Based on a hybrid constellation navigation satellite system, this thesis laid much emphasis on differential positioning and integrated navigation among constellations.The accuracy of pseudorange in hybrid constellation navigation satellite system is diverse among satellites, so the method of weight positioning using user range error can get more precise and accurate result. Likewise, reducing the discrepancy of all observations, pseudorange differential has a better degree than coordinate differential in positioning accuracy.Aiming at the prediction problem of differential correction data, the classic methods are always based on the rate of correction data, while this thesis put forward a new one using delay time neural network. The numerical simulation result indicates that the new method dominates old ones.Because visible satellites in integrated navigation system among constellations are too many, one method to select the best subset is computing the inversion of observation matrix but it costs too much time. To solve this problem, this thesis used radical basic function neural network. Comparing to the old ones, the new method costs less time and gets better degree in accuracy. In addition, it does not degrade by correlated noise.The conclusion of this paper can be directly applied to solve some engineering problems, especially in the field of differential navigation and integrated navigation of hybrid constellation regional navigation satellite system.