Research On Detection And Tracking Of Hypersonic Vehicle Based On Random Finite Set

Hypersonic vehicle flight trajectory belongs to a typical nonlinear motion system.Its flight speed is higher,generally more than 5 Mach,and it is a large range of maneuvering hypersonic target which has stealth capabilities.During the flight process,the number of targets may change over time.These make the tracking of the hypersonic vehicle relatively complex.In light of the situation,this paper presents a kind of algorithms which can be used to track the hypersonic vehicle.The main work includes:(1)Model and parameter analysis of hypersonic vehicle.In view of the nonlinear motion system of hypersonic vehicle,we simplify a six degree of freedom motion equation and get a two dimensional motion equation for it.Then,we analysis the influence of the parameters on the flight trajectory.Finally,the final parameter values are proposed.(2)Hypersonic vehicle tracking algorithm.We combine the multi bernoullli filter(CBMem Ber)with the unscented Kalman filter(UKF),and the generalized labeled multi Bernoulli filter(GLMB)with the UKF.Two algorithms are proposed,repectively.Tracking performance of two algorithms for hypersonic vehicle are analyzed through the contrast experiment.The experiments show that the generalized labeled multi Bernoulli filter has a better tracking performance than that of multi Bernoulli filter.(3)Multi sensors tracking algorithm.The single sensor may loss target due to its limited observation range.Considering multiple sensors is feasible method.We propose a model of detection probability for the sensors.Besides,the concept of extended association graph is introduced.A multi-sensor version for the generalized labeled multi Bernoulli filters is proposed.