Research On Application Of Kalman Filter In Vehicle Integrated Navigation

With the rapid development of social economy and science and technology,the number of vehicles on the market increases significantly,at the same time,it brings the problems of traffic jam,traffic accident and so on,and the application of vehicle navigation equipment can effectively alleviate the traffic problems that brought by increasing vehicles.One of the key technologies to research and develop vehicle navigation is vehicle positioning technology,in the positioning module of vehicle integrated navigation system,the vehicle state information of two positioning systems is usually processed by filtering scheme,thus obtaining better vehicle state estimation and then through map matching process to achieve positioning.When the vehicle state of vehicle integrated navigation system is estimated,the nonlinear filtering algorithm which based on Kalman Filtering is usually used.This thesis based on the theory of Kalman Filtering algorithm,expounded the algorithm theory and calculating process of Extended Kalman Filtering,Iterated Extended Kalman Filtering and Unscented Kalman Filtering;analyzed the characteristics of the experimental data,according to the results of data analysis and the method of data acquisition,establish mathematical model of vehicle integrated navigation-Vehicle velocity model(CV model).On the basis of above-mentioned(the mathematical model and the filtering algorithm),adopt C++ development language and MFC framework to realize the algorithm of Extended Kalman Filtering,Iterated Extended Kalman Filtering,and Unscented Kalman Filtering in vehicle integrated navigation,the state estimation results and its mean square error are displayed in graphical form by TeeChart GUI widget,Baidu map API is applied to display Baidu map on the WebBrowser GUI widget,and vehicle position before and after filtering is presented as icon on Baidu map.Then,compared and analyzed from three aspects as follows:(1)the mean square errors of the three nonlinear filtering results.(2)the closeness between the result position and vehicle driving road which presented on Baidu map.(3)the storage space occupied by the above three kinds of nonlinear filtering algorithm programs.The analysis results show that the Iiterative Extended Kalman Filtering has better localization accuracy and is more suitable for practical engineering.