Adaptive Particle Filter Approach to Approximate Particle Degeneracy
J. Joseph Ignatious, A. UmaMageswari, Simon Abraham · 2012
Abstract – – The main problem of particle filter in nonlinear state estimation is the particle degeneracy. It can be overcome by Resampling operation. But Resampling operation leads to the problem of sample impoverishment. Therfoer an algorithm named Variance reduction technique is proposed to solve sample impoverishment and degeneration problem. It reduces the variance of the particle weights by selecting an exponential fading factor and this factor can be chosen adaptively and iteratively in terms of the effective particle number. Many improved particle filter algorithms were proposed to solve the degeneracy problem which are seemed to be complex. In this paper an algorithm is presented to show that the idea of Variance reduction technique is feasible to propose a new adaptive filter ing algorithm.