Research On Multi-passive Sonar Tracking Algorithm For Bearing-only Target

As countries in the world gradually pay attention to the sovereignty of the ocean,many researchers have carried out research on ship target tracking technology.Since only the azimuth trajectory of the target can be obtained when a single base station is used to passively track a moving target,it is necessary to use the tracking algorithm to extract the azimuth trajectory of the ship target from the measurement information received by the single base station,and then to The target azimuth trajectory tracked by each base station is used to estimate the position of the target through pure azimuth positioning.Under the above background,this paper has carried out the research of multi-passive sonar tracking algorithm for bearing-only targets.First,the passive multi-target tracking algorithm based on data association is studied.The PDA algorithm suitable for single target tracking is studied,and on this basis,an improved PDA algorithm based on frequency information is introduced.After simulation analysis,the performance of the I-PDA algorithm is better than that of the PDA algorithm;the performance of the I-PDA algorithm is studied for multi-target tracking.JPDA algorithm,aiming at the phenomenon that the JPDA algorithm is prone to tracking errors when the target trajectory distance is close or when the target trajectory is close,the improved I-JPDA algorithm based on frequency information is introduced.After simulation analysis,the I-JPDA algorithm can effectively avoid this phenomenon.The occurrence of;the MHT algorithm suitable for multi-target tracking is studied,and its performance is compared with the JPDA algorithm and the I-JPDA algorithm.The effectiveness of the above algorithm is verified by simulation analysis and sea trial experimental data.Secondly,in view of the large amount of calculation in the data association method and the difficulty of considering the new target,the tracking method based on random finite set is introduced.Aiming at the phenomenon that RFS theory cannot obtain an effective closed solution,the GM-PHD algorithm suitable for multi-target tracking is studied,and its tracking performance is analyzed through OSPA distance.After simulation analysis,this algorithm can effectively track new targets and estimate the number of targets;Aiming at the phenomenon that the track may be interrupted and the track is inconsistent during actual tracking,the correlation method based on the interrupted track is studied.After simulation analysis,this algorithm can effectively splice the track.The effectiveness of the above algorithm is verified by sea trial data.Finally,the azimuth trajectory obtained by multiple stations is purely azimuth positioning.The influence of different station layout methods and direction finding errors on positioning accuracy is studied;the direction finding cross positioning method based on least squares is studied.After simulation analysis,this algorithm can effectively locate the target;for the phenomenon that the least squares method is susceptible to noise,Studied the EKF algorithm;in view of the phenomenon that the EKF algorithm is suitable for a system with a low degree of nonlinearity,the UKF algorithm was studied,and the performance of the EKF algorithm and the UKF algorithm under the same conditions was compared.The effectiveness of the above algorithm is verified by simulation analysis and sea trial data.