Research On Detection And Tracking Algorithm Of Discontinuous Acoustic Emitters With Passive Sonars

In terms of remarkable concealed predomainance,high anti-jamming and the ability to extract the attribute information of emitters,etc,passive sonar systems are becoming widely used in underwater detection and tracking fields nowadays.Since the working mechanism of passive sensing,passive sonars may access the intermittent measurements of the discontinuous acoustic emitter that discontinuously emits certain signals.In modern information warfare,the detection and tracking for discontiunuous emitters face severe challenges,such as attribute loss,performance degradation and tracking failure.Consequently,it is an urgent task to track the discontinuous acoustic emitters and enhance the detection accuracy of passive sonar systems.Aiming at the detection and tracking of discontinuous underwater acoustic emitters in passive sonar systems,this thesis mainly focuses on the researches on unknown discontinuous property,low detection accuracy and incomplete observation.The main contributions are as follows:1.Aiming at the acoustic emitters tracking problem with passive sonars,the mechanism of receiving underwater acoustic signals and models of measurements generation by passive sonars are analysed.Then,in the framework of Bayesian estimation theory,the target tracking method is studied,which provides a theoretical foundation for the research of discontinuous emitter detection and tracking in the subsequent chapters.2.Aiming at the tracking problem of discontinuous emitters,two adaptive update time Kalman filters based detect-before-track are proposed(BAUT-KF-DBT and FAUT-KF-DBT).The related system models for discontinuous property of emitters are established,and the tracking algorithm for discontinuous emitters based Bayesian formula is derived.These mehods can estimate both the kinematic state of the emitter and the discontinuous property of the signal emission.The problems of steady detection and robust tracking for discontinuous emitters are solved.3.Aiming at the problem of low tracking accuracy of discontinuous emitters in low signal-to-noise ratio(SNR)environment,a particle filter based track-before-detect varying period estimation(VPE-PF-TBD)approach is proposed.This algorithm integrates the entire measurement intensity by establishing the discontinuous emitter track-before-detect(TBD)measurement model,which can avoid the detection uncertainty of detect-before-track(DBT)method and enhance the performance of tracking weak emitters.4.Aiming at the problem of incomplete observation of single passive sonar,a multiple sonars particle filter track-before-detect based varying period estimation(MS-VPE-PF-TBD)method is proposed.By combining the incomplete discontinuous observations of multiple passive sonars with a centralized fusion strategy,the complete states of discontinuous emitters and discontinuous property are estimated.Then,the detection performance can be improved greatly compared with single passive sonar system.The above work has been verified through the simulation and real-world data,the results demonstrate the effectiveness of the proposed methods.