Abstract
A new mobile robot localization technique is pre- sented which uses multiple Gaussian hypotheses to rep- resent the probability distribution of the robots location in the environment. Sensor data is assumed to be pro- vided in the form of a Gaussian distribution over the space of robot poses. A tree of hypotheses is built, rep- resenting the possible data association histories for the system. Covariance intersection is used for the fusion o/the Gaussians whenever a data association decision is taken. However, such a tree can grow without bound and so rules are introduced for the elimination o/the least likely hypotheses from the tree and for the proper re-distribution of their probabilities. This technique is applied to a feature-based mobile robot localization scheme and experimental results are given demonstrat- ing the effectiveness of the scheme.

This publication has 6 references indexed in Scilit: