Evolutionary Particle Filter: Re-sampling from the Genetic Algorithm Perspective

The Institute of Electrical and Electronic Engineers Inc
Publication Type:
Conference Proceeding
2005 IEE/RSJ International Conference on Intelligent Robots and Systems, 2005, pp. 2935 - 2940
Issue Date:
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail2005002371.pdf349.18 kB
Adobe PDF
The sample impoverishment problem in particle filters is investigated from the perspective of genetic algorithms and remedies are suggested in this research. Studies are conducted for the number of particles required and the time for impoverishment. It is revealed that the sample impoverishment problem is caused by the re-sampling scheme, which should be avoided, in implementing the particle filter with a finite number of particles. Crossover operators from genetic algorithms are adopted to tackle the finite particle problem by omitting the re-sampling process and re-defining particles during filter iterations. The number of particles required and hence the computation complexity are reduced by adapting his genetic operator. Effectiveness of the proposed approach is demonstrated by simulations for a monobot simultaneous localization and mapping application and experimented using a Pioneer mobile robot.
Please use this identifier to cite or link to this item: