A Research Of Pseudo Linear Kalman Filtering Algorithm Based On Bearing Only Target Tracking

The study of bearing-only tracking has significant research value and practical relevance,with widespread applications in both civil and military fields.This approach involves measuring and receiving only the azimuthal angle of a target using sensors,enabling the estimation of its motion state.However,in realistic environments,this system often encounters challenges,such as interference from complex electromagnetic fields and non-Gaussian and nonlinear noise.While the traditional pseudo-linear Kalman filter can handle the nonlinear problems associated with pure orientation target tracking,it suffers from a serious bias problem and cannot maintain high tracking accuracy in non-Gaussian noise environments.To address these challenges,this thesis presents a comprehensive study of the nonlinear filtering and non-Gaussian noise issues in bearing-only tracking and proposes novel algorithms that improve target tracking accuracy.Specifically,this thesis utilizes the pseudo-linear Kalman filtering algorithm,information entropy theory,and optimization constraint method.Firstly,to counteract the impulse noise prevalent in non-Gaussian environments,a pseudo-linear Kalman filter algorithm based on maximum correlation entropy is proposed in conjunction with the traditional pseudo-linear Kalman filter algorithm.This method exploits the correlation entropy in information entropy theory to measure the correlation between the actual state and the estimated state of the target.Simulation results demonstrate that the algorithm exhibits exceptional anti-pulse performance.Secondly,to overcome the bias problem of pseudo-linear Kalman filtering based on maximum correlation entropy,this thesis conducts bias analysis to compensate for the error caused by pseudo-linear noise.This leads to the development of the pseudo-linear Kalman filtering algorithm based on bias compensation,which demonstrates superior tracking performance under different sizes of non-Gaussian noise environments in simulation experiments.Thirdly,to further improve the accuracy of target tracking,this thesis leverages the target velocity range as a priori information,introduces velocity constrained optimization,and proposes a velocity constrained maximum correlation entropy pseudo-linear Kalman filter algorithm.The root mean square error performance of the algorithm is tested under different parameters and sizes of non-Gaussian noise environments through simulation experiments.Finally,experiments on bearing-only tracking based on radar goniometry are conducted,and the experimental results verify the good filtering performance and practicality of the proposed algorithms.