Journal of Physical Agents, Vol 1, No 1 (2007)

Indoor Fuzzy Self-Localization using Fuzzy Segments

David Herrero, Humberto Martinez

Abstract


The research presented in this paper approaches
the issue of indoor localization using the fuzzy logic framework
for modeling and dealing with the uncertainty of the position
measurements. Fuzzy logic presents properties that make it a
suitable tool to represent and manage the different factors that
affect the measures. This framework allows representing the
perceptions, including their associated uncertainty, using fuzzy
sets and making use of the tools provided by the framework to
manage and operate them. This work uses the fuzzy segment
theory to maintain a coherent local representation around the
robot using multi-sensor fusion based on fuzzy logic, and uses
these fuzzy segments to feed a fuzzy self-localization method,
which is able to deal with the ambiguity in the global localization
problem.


Full Text: PDF