Research On Integrated Navigation Orientation System Of GPS And DR In Vehicles

The appearance of Multi-Sensors technology has promoted greatly the development of In-Vehicle Orientation technology, and it ended the time in while GPS or DR navigation system was used solely to orientate for vehicles. It has been shown that the technology of fusion of GPS and DR may improve commendably the precision of navigation for vehicles.In all kinds of theories of information fusion, Kalman filtering theory has been accepted popularly and has been developed quickly and has been used in comprehensive fields, because it has advantages such as recursion, real-time and its information is stored and calculated easily in computers. The Federal Kalman Filtering used in the field of Integrated Navigation Orientation of GPS and DR has been developed under this background. The main idea of this algorithm is that making GPS and DR as its two subsystems, and they are calculated firstly under their own filtering system then their outcomes are carried to the main filter, where these outcomes multiplied the dynamic distributed factors that come from the operation through obscure integration estimate algorithm and dynamic information distributing factors optimization algorithm become the results of optimization fusion.The standard Kalman filtering has visible disadvantages as its advantages. All of these system equations for state vectors that are demanded of must be linear, at the same time the noise met by them must be white. The field of use of the standard KF is limited greatly for the disadvantage. In the thesis, the filter for GPS subsystem will be selected as KF because of the character of its own system equations and under the complicated working condition of mutual correlative color noise. The mutual correlative color noise will be transformed as a "white" noise for "new" system equations through a series of mathematic transformations. In the same way, the filter for DR subsystem will be selected as Unscented Particle Filter which contains the most advantages that Unscented Kalman Filter and Particle Filter own. Though it has a big disadvantage that the particles may be happen to dry up. The author improves this algorithm to avert this shortcoming appearing in operating and calls it as Improved UPF. After having achieved a series of religious mathematic transformations, the author affords flow-charts of the filtering algorithms of two subsystems and the operating structure-charts of them.At the end of this thesis, the author utilized MATLAB to realize simulation. The aim is to prove the feasibility of Integrated Filtering Algorithm and to indicate the precision of its navigation orientation better than that of any subsystem.