Research On Target Tracking Based On Underwater Sensor Network

With the development of information technology and the increase in emphasis on marine, underwater sensor network is developing rapidly, and people pay more and more attention to underwater sensor network-based target tracking. Passive target tracking based on underwater sensor network was studied and an underwater information processing platform was designed in this paper. Main work of this paper are as follows:Fisrtly, single target tracking was studied, an improved resampling non-augmented UPF(Unscented Particle Filter) method was proposed. Non-augmented UKF is used to produce the proposal distribution of particles, costing far less time than using general augmented UKF to do so. In the process of resampling, the discarded particles and replicated particles is combined rationally to produce new particles, improving the diversity of particles. The author applied this algorithm to passive single-target tracking based on underwater sensor network combining with the model of underwater target.Scondly, the improved resampling non-augmented UPF was contaced to MSJPDA (Multisensor Joint Probabilistic Data Association) to do multi-target tracking.Thirdly, simulation proved that single target tracking in this paper is faster and more accurate than other methods, also multi-target tracking is effective even in the situation of orbits crossover. It is worth mentioning that all the work is done in the prerequisite of3D space.Fourthly, an underwater information processing platform was designed using Lab VIEW. The sending end software of the system completes the generation, analysis and sending of the underwater acoustic signal; the target tracking module of the receiving end was cmpleted in the way of simulation. Foundatal work of underwater target tracking has been completed by this pltaform.