Research On Bearings-only Tracking And Locating Technology For Underwater Multi-target

Passive detection technology of underwater target is widely used because of its strong concealment.As the basis of passive sonar detection,multi-target bearings-only tracking technology and passive locating technology play an important role.Firstly,the bearings-only tracking technology is studied,including the principle and application of Modified Polar-coordinates Extended Kalman Filter and Modified Polar-coordinates Unscented Kalman Filter.Then,the accuracy,convergence speed and stability of the tracking filter are verified by simulation experiments.Also,the influence of bearing measurement error on filtering accuracy,convergence speed and stability is studied.Then the multi-target tracking technology is studied.Multi-target bearings-only tracking algorithm based on Probability Data Association algorithm and Multiple Hypothesis Tracking algorithm is used to track the target.When the target tracks cross,MHT algorithm is used to track targets.When the target track is not crossed,the PDA algorithm is used to track the targets.This paper studies the principle and application of PDA algorithm and MHT algorithm.By designing a multi-target tracking process,the application logic of MHT algorithm and the decision logic of track termination are optimized.The detection false alarm caused by the influence of environment noise,clutter,main lobe width,side lobe and underwater acoustic channel is regarded as clutter.Finally,the ocean trial data is used to verify the situation of target track crossing and non-crossing,target track discontinuity and target track termination.Finally,the technology of multi-station target correlation and passive location is studied.The target correlation algorithm is used to correlate the targets observed by multiple observation stations.Then the location is carried out by the algorithm of multi-station bearing intersection.In this paper,two kinds of target correlation algorithms are used to correlate targets,one is based on power spectrum similarity matching and the other is based on time-frequency similarity matching.Then,the accuracy of multi-station target association and the performance of locating algorithm are verified by simulation experiment and trial data.