The Research On Target Track Estimation Technology Base On Passive Sonar

Passive sonar has important application value because of the silent working mode.In the military field,passive sonar undertakes a series of important tasks such as target detection,localization and tracking.In the context of integrated sonar applications,this paper studies passive sonar target track estimation technology.The main research contents include passive sonar measurement acquisition,target tracking filtering,and distributed track fusion under the multi-array posture.The track estimation takes the measurement acquisition as the information input,and Passive sonar ranging is one of the key technologies.The traditional azimuth intersection passive ranging algorithm is simple and easy to implement in engineering,but it requires accurate azimuth,so it is not suitable for long-range targets.This paper analyzed the relationship between target ranging with LOFAR and spatial spectrum based on waveguide invariance,and constructed a passive ranging algorithm based on acoustic interference structure.Compared with the azimuthal intersection passive ranging algorithm,this algorithm requires low accuracy of azimuth and has robust ranging performance,which is suitable for long-range targets and has important significance for passive ranging technology.Trace filtering is one of the key techniques for target track estimation.Kalman Filter is the classical linear Gaussian system state estimation algorithm.Extended Kalman Filter and Unscented Kalman Filter solve the state estimation problem of nonlinear system.Based on the above contents,this paper constructed a state transfer model based on the "current" statistical model and a nonlinear measurement transfer model for passive sonar based on the conversion of polar and Cartesian coordinates,and researched an adaptive Unscented Kalman filter,and the target speed and course estimation equations are derived based on this algorithm to achieve real-time estimation of the target motion state.Distributed structure has higher robustness,better system scalability and extension than other data fusion structures.Track fusion is a distributed data fusion to further improve the robustness and accuracy of the track estimation.The local state estimation error correlation affects the performance of fusion algorithm,and the covariance weighting algorithm is the optimal fusion algorithm under the known error correlation,but it is usually difficult to estimate the correlation accurately in practical applications.In this paper,the ellipsoidal method was used as the research direction to research the Covariance Intersection algorithm under the unknown error correlation condition.Based on this,the Inverse Covariance Intersection algorithm with higher fusion accuracy was investigated,and the covariance ellipsoidal theory was used to demonstrate the effectiveness of this data fusion algorithm.Finally,this paper constructed a comprehensive evaluation index consisting of position root mean square error,state covariance matrix and Hellinger distance to simulate and verify the fusion algorithm.Finally,the passive ranging and target tracking algorithms was validated by processing the sea trial data,and the results show that the algorithms researched in this paper can effectively accomplish the passive ranging and the real-time estimation of target motion parameters such as target track,speed and course.