Research On High Precision GPS Positioning Relative To A Moving Reference And Velocity Determination

Present research on GPS mainly focuses on ordinary relative positioning and precise point positioning.While for satellite formation and attitude determination,users are more concerned about the relative position between the moving vehicles,which need to do much more research on GPS positioning relative to a moving reference.As for the frequent earthquakes in recent years,monitoring crustal deformation is very important,but there is a saturation of the seismograph and strong-motion when strong earthquake occurs.According to above requirements,the research of this thesis are summarized as follow:1.The basic theory of GPS positioning and the method of measuring data processing are studied.The erroe sources of GPS positioning are analysed in detail and we give the solutions accordingly.We also derive the formula of least square mehod and kalman filter and analyse their advangtages and disadvangtages.2.Based on the theory of GPS positioning,we studied the theory of GPS positioning relative a moving reference.The difference between GPS positioning relative a moving reference and ordinary relative positioning are compared and then mathematical model of high precision GPS positioning relative to a moving reference is constructed.the model is validated by static and real kinematic GPS data.3.The baseline length constraint is introduced for baseline length is already known in some cases and we give the mathematical model accordingly.The example of this thesis shown that the proposed method improved the accuracy to some extent.4.The existed methods of velocity determination are summarized and their advantages and disadvantages are analysed.We employ adjacent time-difference o construct derived doppler observation to obtain high precision velocity.The robust least square method is introduced to resist the effect from small cycle slips and gross errors and the method is validated using static data and coseismic data.5.The real-time displacement can be obtained by integrating real-time velocity,for the existing errors of approximate coordinate,satellite orbits and satellite clock,there exsit systematic bias in obtained displacement.The semi-parametric model is introduced to separate systematic bias to improve the accuracy and the model is validated by the example.