Relational Information Space for Dynamic Systems
Abstract
Cooperative decentralized control of autonomous vehicles continues to be an important research subject for many military applications. Vehicles communicate with each other and exchange information about their relative environment and use that data to develop decentralized, coordinated control policies. Communication is often represented by an graph, and information exchange is modeled by a discrete-time dynamical system, known as the information loop. When vehicles agree on the information state they have reached an information consensus. In this work a topological manifold for representing complex networks of dynamical entities is shown to be an effective way to represent non-entropic measures of information as low dimensional embeddings in a high dimensional lifted space. A 2 dimensional embedding was developed that demonstrates an efficient frontier of graphs that dominate all others in their ability to reject disturbances and converge rapidly to a consensus.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 07, 2011
- Accession Number
- ADA542207
Entities
People
- Robert A. Murphey
Organizations
- Air Force Research Laboratory