Set membership localization and mapping for autonomous navigation

In this paper a set theoretic estimation approach is proposed for dynamic localization problems in the area of mobile robot autonomous navigation. Self-localization of mobile robots is one of the most important problems in long range autonomous navigation. When moving in an unknown environment, the navigator must exploit measurements from exteroceptive sensors to build a map, identify landmarks and, at the same time, localize itself with respect to them. This problem is known as simultaneous localization and mapping (SLAM). Under the hypothesis that the errors affecting all sensor measurements are unknown but bounded, set membership techniques, successfully employed in the robust identification area of research, are exploited to devise procedures for guaranteed estimation of robot and landmarks positions in terms of uncertainty regions. Set approximation is adopted in order to provide efficient recursive algorithms, suitable for on-line implementation.