On sample diversity in Particle Filter based robot SLAM

This article proposed a simple and effective methodology for improving diversity of samples in Particle Filter (PF). The motivation lies in the situation that resampling procedure which aims for amending particle degeneracy always leads to particle depletion and less diversity of particles has severe consequence on the filter estimation accuracy. Various resample approaches have been developed in recent years, including multi-nominal resample, residual resample, stratified resample and systematic resample. All of them, however, will inevitably lead to lose of diversity in scatter of particles because they simply replace lower weighed particles with higher weighed particles. In this paper, we developed practical MCMC solutions for drawing particles for PF. The selections of proposal distribution and convergent chain node are taken into careful considerations. It is revealed from the simulations and real experiment that the proposed resampling method is capable of improving the performance of Particle Filter.

[1]  Petar M. Djuric,et al.  New resampling algorithms for particle filters , 2003, 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03)..

[2]  Fredrik Gustafsson,et al.  On Resampling Algorithms for Particle Filters , 2006, 2006 IEEE Nonlinear Statistical Signal Processing Workshop.

[3]  A. Doucet,et al.  Sequential MCMC for Bayesian model selection , 1999, Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics. SPW-HOS '99.

[4]  Ju He-hua.  Optimal Taxis Resample Algorithm and Computer Simulation , 2011 .

[5]  Frank Dellaert,et al.  MCMC-based particle filtering for tracking a variable number of interacting targets , 2005, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Jun S. Liu,et al.  Sequential Monte Carlo methods for dynamic systems , 1997 .

[7]  András György,et al.  A Markov-Chain Monte Carlo Approach to Simultaneous Localization and Mapping , 2010, AISTATS.

[8]  W. Gilks,et al.  Following a moving target—Monte Carlo inference for dynamic Bayesian models , 2001 .

[9]  G. Kitagawa.  Monte Carlo Filter and Smoother for Non-Gaussian Nonlinear State Space Models , 1996 .

[10]  Sebastian Thrun,et al.  FastSLAM: a factored solution to the simultaneous localization and mapping problem , 2002, AAAI/IAAI.

[11]  Nando de Freitas,et al.  Sequential Monte Carlo Methods in Practice , 2001, Statistics for Engineering and Information Science.

[12]  Songmin Jia,et al.  Range scan matching and Particle Filter based mobile robot SLAM , 2010, 2010 IEEE International Conference on Robotics and Biomimetics.

[13]  Christian Musso,et al.  Improving Regularised Particle Filters , 2001, Sequential Monte Carlo Methods in Practice.

[14]  Kevin P. Murphy,et al.  Bayesian Map Learning in Dynamic Environments , 1999, NIPS.

[15]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[16]  Wolfram Burgard,et al.  Exploration with active loop-closing for FastSLAM , 2004, 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (IEEE Cat. No.04CH37566).

[17]  Li Gang.  Estimation of Value-at-Risk Using MCMC , 2000 .

[18]  M. Pitt,et al.  Filtering via Simulation: Auxiliary Particle Filters , 1999 .

[19]  N. Gordon,et al.  Novel approach to nonlinear/non-Gaussian Bayesian state estimation , 1993 .