Swarming in Two and Three Dimensions
Abstract
The future of military action will increasingly require new methods based on 'swarming' tactics in which a multitude of small units or 'pods' can operate in clusters with an overlaying network transmitting information. This project aims at developing concise spatio-temporal models of the large scale dynamics of swarm. The focus is on 'fluid-like' swarms in which the individual units have fairly distributed but localized density. The models have some connection to classical problems in fluid dynamics, with a potential for a richer structure arising from cooperativity between units in the swarm and from self-propulsion of individual units. The justification of the models is based on the internal dynamics swarming as opposed to classical principles of physical fluid flow. Models will be tested against numerical particle-based (Lagrangian) simulations and will be compared with known behavior from biological swarms such as locusts, ants, and fish. This biology-based portion of this research project will include collaboration with Mark Lewis, the Canada Research Chair of Mathematical Biology at the Univ. of Alberta. The second part of this program involves bio-engineering motivated' design of swarm". We consider the inverse problem: given a large scale dynamics for a swarm, how can one design individual motion to achieve this outcome? Our approach is to start with continuum models designed to have desired solutions. We will use knowledge gained from the biological models to derive swarming algorithms that could have both military and industrial use. The designed swarms will include and additional component not present in biological models, that of a communications network distributed among the swarmer subgroups that will facilitate operations.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 28, 2005
- Accession Number
- ADA435127
Entities
People
- Andrea Bertozzi
Organizations
- Duke University