Improved Techniques for the Rao-Blackwellized Particle Filters SLAM

Huan Wang,Hongyun Liu,Hehua Ju,Xiuzhi Li

DOI: https://doi.org/10.1007/978-3-642-10817-4_20

2009-01-01

Abstract:Rao-Blackwellized particle filters simultaneous localization and mapping can yield effective results but it has the tendency to become inconsistent. To ensure consistency, a methodology of an unscented Kalman filter and Markov Chain Monte Carlo resampling are incorporated. More accurate nonlinear mean and variance of the proposal distribution are obtained without the linearization procedure in extended Kalman filter. Furthermore, the particle impoverishment induced by resampling is averted after the resample move step. Thus particles are less susceptible to degeneracies. The algorithms are evaluated on accuracy and consistency using computer simulation. Experimental results illustrate the advantages of our methods over previous approaches.