The Technology Reseach Of Localization And Tracking Of Weak Signal

Localization and tracking of weak signal are widely used in military and civil fields. Because of the signal strength of target weakly, traditional method is not available to detect and track the weak signal. Track-before-detect(TBD) algorithm is based on the raw observations and establishing the tracking model before detecting. The signal energy is accumulated for a specified period of time, then detecting the target sentence, at the same time, the tracking results are output. This method can solve the problem of detection and tracking of weak target in better performance. Particle filter is an effective method to realize the TBD algorithm. This dissertation mainly consider the perspective of noncooperation, and use particle filter to research the detection and tracking of maneuvering weak target using the passive sensor based on angle measurement. The main research results are as follows:Firstly, the basic theory and method of recursive Bayes estimation and particle filter are introduced. The track-before-detect processing model is established based on passive angle measurement sensor. According to the model, the theory of track-before-detect based on particle filter is deeply researched. The unnormalized weight particle filter TBD is established according to the theory. The dissertation analysis and verify the PF-TBD according to the simulation. What’s more, different factors which may influence the performance of PF-TBD are analyzed.Secondly, multiple models particle filter is proposed for the detecting and tracking of the maneuvering weak target. Every particle choose motion model randomly by the rule, and the system choose the particles which are fit the true model using resampling. The target state is gotten by the particles fusion. The location and tracking of maneuvering weak target is partly solved. And the UKF is added to the MMPF-TBD to improve the performance of the algorithm.Thirdly, concerned with the multiple models, the design rule of model set is researched. An improved method by estimating the turn rate is proposed to solve the weakness of multiple models PF-TBD when the target is strongly maneuvering. The turn rate is estimated by filtering, and it is applied to design the model set as parameters. Fewer models get better effect. The practicality of multiple PF-TBD is greatly improved.