Adaptive Double-Resampling Particle Filter Algorithm for Target Tracking

摘要

Based on the traditional particle degradation and depleted of particle filter and the number of particle set, which cannot be adaptive to change brought by the filtering accuracy and convergence rate of decline. A new methods of Innovation and resampling particle filter was applied to the paper. This approach can solve the problems mentioned above. The algorithm first uses the observation information to establish the particle distribution program of the resampling. Then to conduct a resampling on the basis of the initial resampling. The second resampling used the particle cross aggregation algorithm. This can improve efficiency of the particles, and avoid the increase of the calculation when using too many particles. The simulation result based on the DR/GPS shows that compared with the traditional PF algorithm, the algorithm can improve the accuracy and stability of the filter.