Improved UKF And Its Application In Vehicle Navigation In MIMU/BDS

With the rapid development of Beidou project and micro-mechanical inertial devices become gradually maturing,BeiDou Satellite Navigation System will unite with cloud computing and large data resources to build an integration platform,combined with inertial sensors on the terminal form integrated navigation service products,realize intelligent recognition,positioning,tracking,monitoring and management.In the military field,MIMU/ Beidou integrated navigation system can break through the limit of high precision inertial navigation cost,can be widely used for unmanned vehicles,weapons and equipment transport vehicles and troops command vehicle.In civilian areas,with the market of electronic consumer products enlarge,mobile phone and land vehicle navigation as the representative.Beidou chip and silicon micro sensor combination technology will contribute to the realization of networking,smart city in china.On the basis of low cost to improve the precision and stability navigation system is one of the main subject in this research.In this paper,with the combined filter as the breakthrough point to research,main content as below:Firstly analyzes the Status and Trend integrated navigation system base on MIMU,the study of mechanism based on and the main research direction of several brief analysis.Then the thesis described the principle of strapdown inertial navigation system and positioning and velocity measurement of Beidou receiver.Next established system model of direct navigation,and provides a basis for the follow-up study of information fusion filtering.Study on the solution of random error and uncertainty error,analysis power spectrum of the MIMU original data,establishment of sensor error model.Alignment method of strapdown inertial navigation system suitable for low precision of IMU,proposes a dual vector attitude determination based accelerometers/magnetometry combined coarse alignment method.In order to satisfy the robustness and address the nonlocal sampling problem inherent in which system dimension mobile condition is too high.Research on the improvement of UKF filter in MIMU system.Aiming at requirements and characteristics of the low precision inertial system,using the method of amplify stability domain filtering by fading factor.Derivation and simulation of strong tracking filter based on UT transform.Moreover,this paper derivate TUKF algorithm that solve the nonlocal sampling problem.In this paper,writer analyzed the performance with simulation to improve the high dimensional system.Proposed combination of strong tracking UKF transform sampling points based on the two algorithms(ST-TUKF),and gives the detailed steps of the improved algorithm.Research on the problem of outage of Beidou signal,combined with the smoothing algorithm using the whole navigation section data to dead reckoning navigation parameters of BD failure period,we derive Two Filter Smoothing algorithm,designed the system switching model,implementation of the bridge failure time navigation information.Application of UKF in the combined system,separately to the laboratory static test and vehicle test data for off-line processing analysis,test results by processing error comparison navigation attitude determination and positioning of the two groups were analyzed and the reasons of error and improve the precision of the method are discussed.And simulation verified the effectiveness of the proposed methods.