| INS and GPS are the most widely used navigation systems currently,but due to the inevitable shortcomings they exist,if used independently,none of them are able to meet the increasingly harsh requirements of the navigation performance.However,if they can be combined by some means,through which their strengths can make up for each other's shortcomings,the system's accuracy and reliability will be greatly improved.Data fusion is the key technology to realize the integrated navigation system,and the key of data fusion is the processing of filtering algorithm,so it is very important to the design and improve the filter in the integrated navigation system.Unscented Kalman Filter(UKF)is a kind of non-linear filtering method which has been proposed in recent years and can be applied in any nonlinear non-Gaussian system without first-order linearization of nonlinear system functions,Making the system less affected by the linearization error.Therefore,since its date,it has been in the project has been more widely used.This paper mainly focuses on the improvement of UKF algorithm and its application in SINS/GPS integrated navigation system.The main contents of this paper are as follows:Firstly,several commonly used coordinate system and the transformation between the coordinate system in the strapdown inertial navigation system and several key model parameters in the Earth reference model are introduced.Then,derived the mechanical layout equation of the strapdown inertial navigation system,including the basic equations of inertial navigation system and the speed,position,attitude calculation equation,and this basis,the nonlinear error model of the strapdown inertial navigation system is deduced.Briefly introduces the working principle of GPS system and analyzes the main errors in GPS navigation and positioning process.Otherwise,the way about combination of SINS and GPS is simply introduced,and the structure diagram of each combination mode is given.According to the error analysis,the SINS/GPS loosely coupled and tightly coupled systems are modeled separately,and the equation of state and measurement are given.Then,the UKF algorithm is introduced in detail,and the core mathematics foundation which named UT transform of UKF is studied,The comparison between the symmetric sampling and the minimum skewness singular sampling strategy is analyzed.The computational complexity of the combined navigation filter algorithm is reduced by using the minimum skewness singular sampling,and the proportional sampling correction algorithm commonly used in symmetric sampling is introduced.Degree of single-shaped sampling,toimprove the accuracy of UKF filter,analyze the effect of the scaling factor in the UT transformation,a kind of selection strategy of adaptive scale factor is proposed and proposes this kind of adaptive scaling factor algorithm embedded in the simplex sampling UT transform in.Then,a simple six-dimensional nonlinear model is used to simulate comparison the three UKF algorithm,which are symmetrical sampling UKF,minimum skewness simplex sampling UKF,and single-sample UKF algorithm with adaptive proportional parameter,to verify the performance of the improved UKF algorithm.Finally,according to the characteristics of the SINS/GPS integrated navigation system model,the adaptive proportional parameter? and the simplex sampling UKF are applied to the integrated navigation system,aiming at solving the problems of low precision and high computational cost.Apply the three UKF algorithms,which are symmetrical sampling UKF,minimum skewness simplex sampling UKF,and single-sample UKF algorithm with adaptive proportional parameter,to SINS/GPS loosely coupled and tightly coupled systems respectively,and the simulation results of these three UKF algorithms are compared and analyzed.In the last of this paper,apply the adaptive proportional parameter UKF to real vehicle navigation data processing,and verify the performance of the algorithm,and compared with the traditional symmetric UKF algorithm. |
PDF Full Text Request