Adaptive formation control for non-holonomic multi-agent systems
(English)Manuscript (preprint) (Other (popular science, discussion, etc.))
The current paper treats the problem of formation keeping for a non-holonomic multi-agent system. In the paper it is shown that tracking controls developed for single-agent systems can be used as base functions for multi-agent formations. To get an idea of the performance that can be expected for such formations, robustness to noise and error propagation are studied in simulations for a line formation with simple internal structure. In addition, it is shown that switching between different tracking angles and tracking controls can safely be performed. The results on switching stability imply that it is possible to change the structure of a formation on-line. Together, the results obtained in the paper open up for a variety of applications where a team of robots cooperate to perform tasks faster and with higher precision than would be possible for a single agent. To illustrate the potential of the results, two case studies are included in the paper. They are both independent of the rest of the paper and are added as an Appendix.
IdentifiersURN: urn:nbn:se:kth:diva-10906OAI: oai:DiVA.org:kth-10906DiVA: diva2:231579
QC 201007152009-08-142009-08-142010-07-15Bibliographically approved