Unified Fusion Approach with Application to SLAM

In this paper, we propose the pre-processor based on the Evidence Supporting Measure of Similarity (ESMS) filter and also propose the unified fusion approach (UFA) based on the general fusion machine coupled with ESMS filter, which improve the correctness and precision of information fusion in any fields of application. Here we mainly apply the new approach to Simultaneous Localization And Mapping (SLAM) of Pioneer II mobile robots. A simulation experiment was performed, where an autonomous virtual mobile robot with sonar sensors evolves in a virtual world map with obstacles. By comparing the result of building map according to the general fusion machine (here DSmT-based fusing machine and PCR5-based conflict redistributor considereded) coupling with ESMS filter and without ESMS filter, it shows the benefit of the selection of the sources as a prerequisite for improvement of the information fusion, and also testifies the superiority of the UFA in dealing with SLAM.

Authors:

• Xinde Li
• Xinhan Huang
• Min Wang

Keywords:

• DSmT
• ESMS filter
• SLAM
• UFA

References:

[1] U. Frese, P. Larsson, T. Duckett (2005). A Multilevel Relaxation
Algorithm for Simultaneous Localization And Mapping. IEEE Trans.
Robotics and Automation, 21(2), 196-207.
[2] Shafer G. A. (1976). Mathematical Theory of Evidence. Princeton
University Press, Princeton, NJ, 297pp.
[3] Ph. Smets (2000). Data Fusion in the Transferable Belief Model,
Presented at the 2000 Int. Conf. Information Fusion, Paris.
[4] Smarandache F., Dezert J. (Editors)( 2004), Advances and Applications
of DSmT for Information Fusion, American Research Press, Rehoboth, ,
Available: http://www.gallup.unm.edu/~smarandache/DSmT-book1.pdf
[5] F. Smarandache (2004) Unification of Fusion Theories (UFT).
International Journal of Applied Mathematics & Statistics, vol. 2,1-14.
[6] R.R.Yager(1983). Hedging in the Combination of Evidence. Journal of
Information and Optimization Science, 4(1), 73~81.
[7] D. Dubois, H. Prade (1988). Representation and Combination of
Uncertainty with Belief Functions and Possibility Measures.
Computational Intelligence, vol.4, 244~264.
[8] M. Daniel (2000). Distribution of Contradictive Belief Masses in
Combination of Belief Functions. Information, Uncertainty and Fusion,
Kluwer Academic Publishers, 431~446
[9] F. Smarandache, J. Dezert (2005). Information Fusion Based on New
Proportional Conflict Redistribution Rules. Presented at the 2005 Int.
Conf. Information Fusion, Philadelphia, PA, USA.
[10] X. Li, J. Dezert, X. Huang (2006). Selection of Sources as a Prerequesite
for Information Fusion with Application to SLAM. Proc. Int. Conf. on
Information Fusion, Florence, Italy, July 10-13.
[11] J. Dezert, F. Smarandache (2003). On the Generation of Hyper-power
Sets for the DSmT. Presented at the 2003 Int. Conf. Information Fusion,
pp.1118-1125, Cairns, Queensland, Australia,8-11 .
[12] J. Dezert, F. Smarandache (2003). Partial Ordering of Hyper-power Sets
and Matrix Representation of Belief Functions within DSmT. Presented at
the 2003 Int. Conf. Information Fusion, pp.1230-1138, Cairns,
Queensland, Australia, 8-11.
[13] B. Ristic, Ph. Smets (2006) Association of Uncertain Combat ID
Declarations. Presented at the Proc. of Cogis-06 Conf.,Paris.
[14] X. Li, X. Huang, M. Wang (2006). A Comparision of The Effect of Sonar
Grid Map Building Based on DSmT and DST. IEEE Int Conf
WCICA2006, Dalian, China, in press.
[15] X. Li, X. Huang, M. Wang (2006). Robot Map Building from Sonar
Sensors and DSmT," Inf. & Sec.: An Int. J., Bulg. Acad. of Sci., Sofia.
[16] A. Martin, C. Osswald (2006). A New Generalization of The
Proportional Conflict Redistribution Rule Stable in Terms of Decision. To
appear in "Advances and Applications of DSmT for Information Fusion",
Vol.2, July 2006, F. Smarandache and J. Dezert (Editors), in preparation.