Research On UUV Estimation Method Of Tracking Target Motion Elements

At present,China’s marine industry has entered a good stage of development,and the rapid development of science and technology today is the best period for China to vigorously build a strong marine nation.Unmanned Underwater Vehicle(UUV)is currently a very popular economic and safe underwater equipment,is the best choice for performing marine missions.Underwater target motion element estimation is one of the core technologies of UUV to realize underwater target detection.In order to improve target motion element estimation performance,this paper focuses on nonlinear non-Gaussian noise filtering algorithm based on target motion element estimation,UUV target tracking path optimization algorithm In-depth research has been carried out in two aspects.The main work of the paper is:First,the development status of underwater target motion element estimation and filtering theory is briefly described,and the core problem of underwater target motion element estimation is clarified.Establish commonly used underwater target tracking models,give target observable conditions through analysis.It also elaborates the identification method of target motion elements based on Bayesian theory,which lays the foundation for the subsequent research.Second,in order to select the optimal filtering algorithm to estimate the target motion elements,the nonlinear Kalman filter is studied,and the Extended Kalman Filter(EKF),Unscented Kalman Filter(UKF),and Cubature Kalman Filter(CKF)are analyzed and compared,and the superiority of CKF in the target tracking model is determined.Due to the linearity of the target tracking state equation,based on Square Root Cubature Kalman Filter(SCKF),the idea of edge sampling is introduced to construct the Marginal Square Root Cubature Kalman Filter(MSCKF)The linear mapping of the state matrix is used to calculate the one-step predicted state and the one-step predicted state covariance matrix.Furthermore,in order to solve the problem that the a priori information of target motion is difficult to acquire and the statistical characteristics of process noise are unknown,an Adaptive Marginal Square Root Cubature Kalman Filter(AMSCKF)filtering algorithm is proposed.Sage-Husa timevarying noise statistical estimator and edge square root Kalman filter MSCKF use state matrix and measurement information to estimate the process noise covariance matrix in real time,and verify the effectiveness of the filtering algorithm through simulation.Then,for the problem that Sage-Husa cannot guarantee the positive definiteness of the estimated process noise covariance matrix,it is easy to cause the square root filter recursion to fail,and it is easy to diverge and the amount of calculation is large,the Strong Tracking Filter(STF)is introduced to construct MSCKF-STF Filter.According to the orthogonality principle,by introducing time-varying fading factors to modify the state prediction error covariance matrix in real time,and then adjust the filter gain matrix in real time,improving the filter’s robustness to model uncertainty Stability and ability to track mutations.Further,for the problem of non-Gaussian noise in the target tracking system under the interference of sea clutter and compound K noise,MSCKF-STF is applied to the Particle Filter(PF)framework to construct the MSCPF-STF filter,using real-time observation data,through MSCKF-STF optimizes the particle importance density function to improve the filtering accuracy while taking into account the non-Gaussian noise characteristics.In order to more accurately describe the complex motion process of tracking targets,and apply MSCPF-STF to the Interacting Multiple Model(IMM)filtering framework,construct the IMM-MSCPF-STF filter,and verified the effectiveness of the filtering algorithm by simulation.Finally,in order to further improve the estimation accuracy of target motion elements,the influence of UUV maneuvering route on the estimation accuracy of target motion elements is analyzed,and the UUV trajectory optimization method based on the maximum azimuth angle change rate is introduced.For the trajectory optimization problem,the UUV trajectory optimization method with the largest azimuth change rate is combined with the multi-model IMM-MSCKF-STF.Based on the model probability,the UUV optimal maneuver route calculated by each filter is fused interactively.Finally,the effectiveness of the algorithm is verified through simulation comparison.