Research On Tracking Uniform Linear Moving Target Based On Kalman Filter

There are a lot of nonlinear systems in the engineering using dynamic systems.When applied to practice,it should be studied according to the different analysis methods and system descriptions.Both linear system and nonlinear system need a filtering algorithm which is suitable for their own real conditions to accurately estimate and predict their state changes.In many engineering application fields of nonlinear system,such as automatic control,weapon guidance,moving target tracking,three-dimensional navigation and so on,high precision is needed to achieve the system purpose,which puts forward very high requirements for filtering algorithm.In this paper,a two-dimensional uniform motion nonlinear system model is selected,and its mathematical modeling and simulation analysis are carried out under two Kalman filtering algorithms.The research contents and results are as follows:1.According to the realization way of the extended Kalman filter algorithm,the paper expounds three mathematical concepts,which are Taylor series,Jacobian matrix and Hessian matrix,analyzes and compares the Kalman filter dynamic system of linear model and nonlinear model in state space,and describes the realization steps of the extended Kalman filter algorithm.At the same time,the idea of Unscented Kalman filter algorithm is introduced,and the UT transformation and proportional correction sampling strategy used in the algorithm are shown.At the same time,the implementation process of Unscented Kalman filter algorithm is expanded.2.The extended Kalman filter algorithm is logically analyzed,and the system's propulsion process is identified by graph.The two core contents of the algorithm,Taylor linearization and Jacobian matrix replacement,are analyzed.A mathematical model based on the extended Kalman filter algorithm is established for a two-dimensional nonlinear system.The simulation software MATLAB is used to calculate the simulation results.It is found that it has a good tracking and estimation effect on the direction and distance of the target in the twodimensional uniform motion.The tracking estimation error is not as large as the observation error,but there will be large filtering error in some segments.3.The algorithm of Unscented Kalman filter is analyzed logically,and the parameter selection and weight calculation are carried out.A mathematical model based on the unscented Kalman filter algorithm is established for the above-mentioned nonlinear system.The simulation software MATLAB is used to calculate the simulation results.It is found that in the two-dimensional uniform motion under the algorithm,the tracking estimation effect of azimuth and distance is also good,and the simulation results can show that the algorithm is superior to the extended Kalman filter algorithm under the mathematical model.4.The distribution of sampling points in UT transformation is transformed.The results show that although the error is small,it may bring filtering divergence,so it is not suitable to use UKF algorithm.