The Target Tracking Algorithm Research Based On The Particle Filtering

The fast advances of computers in the last several years made particle filtering, as a powerful method for sequential signal processing, a very active area of research. It has captured the attention of many researchers in various communities including those of signal processing, statistics, and econometrics, and this interest stem from its potential for coping with difficult nonlinear and/or non-Gaussian problems. In particular the concern of national defense, because in the target tracking research field it has become effective.The problems of several variants of filtering such as EKF, Grid-Based Method, EKF and Approximate Grid-Based Method are discussed. Aiming at the nonlinear and/or non-Gaussian filter problems, the generic ideas of particle filter are given, based on the analysis of standard algorithm of sampling-importance-resampling filter, the problems of particle filter are discussed and some improvement methods are illustrated. Pseudo-code of every variants of particle filter is given. Several variants of particle filter such as SIR,ASIR,RPF are compared. The advantages and disadvantages of them are discussed.The prediction sample propagated by general PF do not consider the last state value, so compared sample by this algorithm with the true posterior distribution the deviation is greater. It is certain blindness. Especially when prognosticated data appear at the coda of transferred probability distribution or the likelihood function is far tight (the perk), general particle filter may be expired. So by choosing and sampling importance dentist based on the likelihood function, a new algorithm is proposed. At last ,in a simulation comparison with EK,FPF and the new algorithm, it is proved the new algorithm yield better performance than EKF and PF in target tracking.