A New MCMC Particle Filter: Re-sampling Form the Layered Transacting MCMC Algorithm

Abstract

In this paper, a new method, named layered trans- acting MCMC Resampling algorithm, is proposed to handle the sample impoverishment problem. The basic idea of the new method is to adjust all particles to the high likelihood areas in state-space rather than multiplying particles with high weights and eliminating particles with small weights, which avoids sample impoverishment effectively. In the proposed method, mutation operator and Particle Swarm Optimization (PSO), which considered as transition kernels of MCMC, applied to each particle, and this promotes a possible displacement of the particles to a better location in the state-space until converging to target posterior density. Finally,a computer simulation is performed to show the effectiveness of the proposed method. Keywords-particle filter; mutation; PSO; Markov Chain Monte Carlo; sample impoverishment;resampling