A hierarchical approach to the multi-vehicle slam problem
Date of Issue2012
International Conference on Information Fusion (FUSION) (15th : 2012)
School of Electrical and Electronic Engineering
In this paper we present a novel hierarchical solution to the Multi-Vehicle SLAM (MVSLAM) problem by extending the recently developed random finite set (RFS) based SLAM filter framework. Instead of fusing control and measurement data at each time step, we introduce a RFS Single-Vehicle SLAM based sub-mapping process, where each robot periodically produces a local sub-map of its vicinity and communicates the resultant sub-map along with the sequence of applied control commands for further fusion into a higher level MVSLAM algorithm, reducing the required network bandwidth and computational power at the fusion node. Our solution is based on the factorization of MVSLAM posterior into a product of the vehicle trajectories posterior and the landmark map posterior conditioned on the vehicle trajectory. The landmarks and the measurements are modelled as RFSs, instead of using data from exteroceptive sensors, measurements are derived from the produced local sub-maps. The vehicle trajectories posterior is estimated using a Rao-Blackwellised particle filter, while the landmark map posterior is estimated using a Gaussian mixture (GM) probability hypothesis density (PHD) filter.
DRNTU::Engineering::Electrical and electronic engineering
© 2012 ISIF. This paper was published in [2012 15th International Conference onInformation Fusion (FUSION)] and is made available as an electronic reprint (preprint) with permission of ISIF. The paper can be found at the following official DOI: [http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=6289934&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D6289934]. One print or electronic copy may be made for personal use only. Systematic or multiple reproduction, distribution to multiple locations via electronic or other means, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper is prohibited and is subject to penalties under law.