Doktorarbeit: Kartierung und Lokalisation durch mobile Serviceroboter unter der Annahme unbekannter aber begrenzter Sensorfehler

Kartierung und Lokalisation durch mobile Serviceroboter unter der Annahme unbekannter aber begrenzter Sensorfehler

Buch beschaffeneBook-Anfrage

Schriftenreihe technische Forschungsergebnisse, volume 22

Hamburg , 190 pages

ISBN 978-3-8300-8397-9 (print) |ISBN 978-3-339-08397-5 (eBook)

About this book deutschenglish

Mobile service robots, equipped with sensors for environmental perception, can be applied in a multitude of tasks. Due to technical reasons, sensor data is always influenced by some kind of error. Hence, approaches for the navigationof mobile service robots have to make certain assumptions about these sensor errors. The aim of the presented work is the investigation of a previously rather unknown approach for the modelling of sensor errors, and its application in algorithms for mapping and localization with a mobile service robot. In contrast to well-known probabilistic approaches, the sensor models are developed by means of interval arithmetic, or interval analysis. These models assume that outer boundaries for the sensor errors can always be stated. They are known as unknown but bounded sensor errors, because the exact characteristics of the error distribution inside the boundaries is unknown. First, two map models are developed for mapping. Both are map models of variable grid size, allowing an efficient and dynamic spatial separation. Both can be represented by hierarchical data structures. A mapping approach will be developed based on the new sensor models for the calculation of these environmental map models. To allow for computation of incremental and global localization of a mobile service robot, two algorithms are presented. Both are based on the presented interval-based sensor models and can be executed on the new map models as well as on a simple line map. For the evaluation of the developed appoaches, they will be investigated by simulation and in real indoor environments. A comparison to their probabilistic counterparts is presented.

Ihr Werk im Verlag Dr. Kovač

Bibliothek, Bücher, Monitore

Möchten Sie Ihre wissenschaftliche Arbeit publizieren? Erfahren Sie mehr über unsere günstigen Konditionen und unseren Service für Autorinnen und Autoren.