Open Collections

Autonomous and cooperative multi-robot system for multi-object transportation

University of British Columbia

This thesis investigates multi-robot cooperation in multi-robot systems (MRS) for simultaneous multi-object transportation in unknown, dynamic, and unstructured environments. Two distinct control frameworks are developed for MRS to achieve its global goal while resolving conflicts in the system. An autonomous and distributed algorithm that uses artificial immune system (AIS) is developed for multi-robot cooperation and it is validated using experimental work carried out on a team of real physical robots in the Industrial Automation Laboratory (IAL). Two deadlock handling approaches are considered to avoid shared resource conflicts in the system. One method prevents the system from getting into a deadlock situation and is so-called the prevention-based method. The other method autonomously detects the deadlocks and then recovers system from that situation, and is known as the detection-based method. Two separate deadlock resolution algorithms are developed for MRS; one based on the prevention method and the other based on the detection method. Either of these two deadlock handling algorithms is then combined with the multi-robot cooperation algorithm to generate two integrated task execution algorithms for the control frameworks of MRS. Feasibility and effectiveness of the developed control frameworks are demonstrated and evaluated through simulation of the MRS on the Webots simulation platform. Finally, using the simulation results, a comparative evaluation of the two control frameworks developed in this research is carried out with respect to task completion time, communication overhead, and the number of tasks executed.

Text

2014-09-10

Attribution-NonCommercial-NoDerivs 2.5 Canada

http://hdl.handle.net/2429/50337

Mechanical Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/2.5/ca/