Η αυτόνομη λειτουργία των ρομπότ εντός περίπλοκων χώρων εργασίας αποτελεί ένα επίκαιρο θέμα έρευνας και η αυτόνομη πλοήγηση είναι αναμφισβήτητα ένα θεμελιώδες κομμάτι αυτής. Επιπλέον, καθώς οι εργασίες που τα ρομπότ καλούνται να εκπληρώσουν αυξάνονται σε πολυπλοκότητα μέρα με τη μέρα, η χρήση πολύ-ρομποτικών συστημάτων, τα οποία εμφανίζουν γενικά υψηλότερη ευρωστία και ευελιξία, αυξάνεται προοδευτικά. Ως εκ τούτου, τα προβλήματα αυτόνομης πλοήγησης που πρέπει να επιλυθούν γίνονται όλο και πιο απαιτητικά, αυξάνοντας την ανάγκη για πιο αποτελεσματικά και σθεναρά σχήματα σχεδιασμού πορείας και κίνησης.
(EL)
The autonomous operation of robots inside obstacle cluttered environments constitutes an actively studied research topic and autonomous navigation is undeniably a fundamental aspect of it. Moreover, as the tasks that robots are called to fulfill grow in complexity by the day, the employment of multi-agent robotic systems, which generally exhibit higher robustness and versatility than their single-robot counterparts, progressively increases. Thus, the autonomous navigation problems that need to be addressed become more and more challenging, increasing the need for more efficient and robust path and motion planning schemes.
In this dissertation, we employ hierarchical decomposition techniques, coupled with suitably designed adaptive configuration space exploration schemes for building versatile and complete planners, capable of computing "paths" connecting two given configurations, or determine the absence of feasible solutions, after a finite amount of steps. In addition, distributed control laws based on artificial potential fields, which can elegantly realize the generated high-level plans, are devised for addressing the navigation problem within complex, planar workspaces with guaranteed convergence properties. Finally, we demonstrate the efficacy of the proposed methodologies for addressing real-world problems, such as coordinating the motion of multiple robots operating within the same environment, as well as coordinating the transportation of an object by a team of cooperating mobile manipulators.
(EN)
National Technical University of Athens
