Complexity study of guaranteed state estimation applied to robot localization
Emmanuel Seignez, Alain Lambert · 2008
The estimation of a vehicle configuration in its environment is mostly solved by Bayesian methods. Interval analysis allows an alternative approach: bounded-error localization. Such an approach provides a bounded set of configuration that is guaranteed to surround the actual vehicle configuration. This method is validated in real time through experimental results on a mobile robot in an indoor environment.