Maximilian Ernestus and Dominik Krupke
Abstract: In this bachelor thesis, we examine a scalable and distributed method to control and maintain large swarms of robots. In an attempt to find a balance between local and global swarm manipulation strategies, independent, externally controlled leader robots are used to influence the shape and movement of a swarm. Particular focus is put on the distributed management of swarm connectivity in the face of multiple simultaneous robot failures. Steiner-tree approximations are formed between the leader robots to pursue the antagonistic goals of control and connectivity. A simulator is developed to allow for simulated experiments to verify the feasibility of the proposed methods and an intuitive understanding of the swarm algorithms by its extensive visual analysis capabilities.
Follow-up paper
- Maximilian Ernestus, Sándor P Fekete, Michael Hemmer, and Dominik Krupke. Continuous Geometric Algorithms for Robot Swarms with Multiple Leaders. In European Workshop on Computational Geometry (EuroCG 2015), 2015.
- Dominik Krupke, Maximilian Ernestus, Michael Hemmer, and Sándor P Fekete. Distributed cohesive control for robot swarms: Maintaining good connectivity in the presence of exterior forces. In Intelligent Robots and Systems (IROS), 2015 IEEE/RSJ International Conference on, pages 413--420. IEEE, 2015.
