This chapter presents a reliable method to detect and verify the existence of loops along the uncertain trajectory of a robot, based on proprioceptive measurements only, using a bounded‐error approach. In a reliable context, a distinction must be made between the detection and the verification of a loop. Indeed, the complexity of SLAM algorithms quickly increases with the exploration of wide environments, as it implies several loop closures to be identified among a dense set of data. The topological degree theory will be used for proving the existence of loops. The chapter aims to underline the usefulness of the degree test for zero detection of functions with bounded uncertainty, and its relevance for loop closure proofs. The efficiency of the new method has been demonstrated on actual experiments involving the AUVs Redermor and Daurade. A total of 25 complete loop‐detection sets have been computed on this test‐case, with the other solutions being partial.
Looped Trajectories: From Detections to Proofs
Reliable Robot Localization ; 133-163
2019-12-16
31 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
A single channel looped digital optical processor
British Library Conference Proceedings | 1990
|NTRS | 2020
|