Research On Symmetry Measurement Equation Based On Multiple Extended Target Tracking

Multi-target tracking technology is one of the important research topics in the field of information fusion.It has wide application prospects in the military and civilian fields and is highly valued by researchers and engineers at home and abroad.This dissertation mainly studies the multi-extended target tracking algorithm based on symmetric measurement equations(SME).The main achievements are as follows:The SME filtering algorithm based on the quadrature Kalman filter and uncorrelated conversion filter is proposed to solve the non-linearity of the measurement equation brought by the SME symmetry function.The proposed algorithms uses the SME symmetry function to transform the measurement equation,and then uses a nonlinear filter to estimate the target state.Simulations show that the proposed algorithm can effectively track the target,correctly distinguish different targets and track the trajectory of multiple targets.A measurement partition algorithm based on the kernel partition is proposed to solve the problem of target missed estimation under the near-crossing scenario of the extended target.The proposed algorithm divides the measurement into the corresponding state with the maximum likelihood probability,and filters the clutter by setting the threshold value.Simulation show that the kernel partitioning algorithm can effectively filter clutter and solve the problem of missed estimation of neighboring scenes.On the basis of the SME method and kernel partition,the SME algorithm is extended to multi-extended target tracking,and Gaussian kernel functions are used as symmetric functions to solve the problem that the excessive number of targets leads to the high degree of nonlinearity of traditional SME symmetric functions.Simulations verify the effectiveness of the proposed algorithm.A probability hypothesis density filter algorithm based on multiplicative noise model is proposed to solve the problem of multi-extended target shape estimation.This proposed algorithm combines the multiplicative noise model and the probability hypothesis density algorithm based on the random finite set framework.Simulation show that the proposed algorithm can effectively estimate the motion state and the extended morphology of the multi-elliptic extended target.For the multi-extended target tracking problem modeled by random matrix,an SME multi-extended target tracking algorithm combined with the random matrix model and the SME algorithm is proposed.Simulations show that the proposed algorithm can effectively estimate the extended target of the motion state and the extended morphology and can correctly distinguish different trajectory of multiple targets.