Klodt, Lukas (2017)
Coordination of Cooperative Multi-Robot Teams.
Technische Universität Darmstadt
Ph.D. Thesis, Primary publication
|
Text
Diss_final.pdf - Accepted Version Copyright Information: CC BY 4.0 International - Creative Commons, Attribution. Download (3MB) | Preview |
Item Type: | Ph.D. Thesis | ||||
---|---|---|---|---|---|
Type of entry: | Primary publication | ||||
Title: | Coordination of Cooperative Multi-Robot Teams | ||||
Language: | English | ||||
Referees: | Adamy, Prof. Dr. Jürgen ; Faigl, Prof. Jan | ||||
Date: | 2017 | ||||
Place of Publication: | Darmstadt | ||||
Date of oral examination: | 30 November 2017 | ||||
Abstract: | This thesis is about cooperation of multiple robots that have a common task they should fulfill, i.e., how multi-robot systems behave in cooperative scenarios. Cooperation is a very important aspect in robotics, because multiple robots can solve a task more quickly or efficiently in many situations. Specific points of interest are, how the effectiveness of the group of robots completing a task can be improved and how the amount of communication and computational requirements can be reduced. The importance of this topic lies in applications like search and rescue scenarios, where time can be a critical factor and a certain robustness and reliability are required. Further the communication can be limited by various factors and operating (multiple) robots can be a highly complicated task. A typical search and rescue mission as considered in this thesis begins with the deployment of the robot team in an unknown or partly known environment. The team can be heterogeneous in the sense that it consists of pairs of air and ground robots that assist each other. The air vehicle – abbreviated as UAV – stays within vision range of the ground vehicle or UGV. Therefrom, it provides sensing information with a camera or similar sensor that might not be available to the UGV due to distance, perspective or occlusion. A new approach to fully use the available movement range is presented and analyzed theoretically and in simulations. The UAV moves according to a dynamic coverage algorithm which is combined with a tracking controller to guarantee the visibility limitation is kept. Since the environment is at least partly unknown, an exploration method is necessary to gather information about the situation and possible targets or areas of interest. Exploring the unknown regions in a short amount of time is solved by approaching points on the frontier between known and unknown territory. To this end, a basic approach for single robot exploration that uses the traveling salesman problem is extended to multirobot exploration. The coordination, which is a central aspect of the cooperative exploration process, is realized with a pairwise optimization procedure. This new algorithm uses minimum spanning trees for cost estimation and is inspired by one of the many multi-robot coordination methods from the related literature. Again, theoretical and simulated as well as statistical analysis are used as methods to evaluate the approach. After the exploration is complete, a map of the environment with possible regions of higher importance is known by the robot team. To stay useful and ready for any further events, the robots now switch to a monitoring state where they spread out to cover the area in an optimal manner. The optimality is measured with a criterion that can be derived into a distributed control law. This leads to splitting of the robots into areas of Voronoi cells where each robot has a maximum distance to other robots and can sense any events within its assigned cell. A new variant of these Voronoi cells is introduced. They are limited by visibility and depend on a delta-contraction of the environment, which leads to automatic collision avoidance. The combination of these two aspects leads to a coverage control algorithm that works in nonconvex environments and has advantageous properties compared to related work. |
||||
Alternative Abstract: |
|
||||
URN: | urn:nbn:de:tuda-tuprints-70265 | ||||
Classification DDC: | 600 Technology, medicine, applied sciences > 620 Engineering and machine engineering | ||||
Divisions: | 18 Department of Electrical Engineering and Information Technology 18 Department of Electrical Engineering and Information Technology > Institut für Automatisierungstechnik und Mechatronik 18 Department of Electrical Engineering and Information Technology > Institut für Automatisierungstechnik und Mechatronik > Control Methods and Robotics (from 01.08.2022 renamed Control Methods and Intelligent Systems) |
||||
Date Deposited: | 15 Dec 2017 11:29 | ||||
Last Modified: | 15 Dec 2017 11:29 | ||||
URI: | https://tuprints.ulb.tu-darmstadt.de/id/eprint/7026 | ||||
PPN: | 42402554X | ||||
Export: |
View Item |