Gyroscopic Many Body Problems in Cooperative and Adversarial Control
Abstract
In this project, the primary aim is to advance the science of controlling collectives of systems engaged in cooperative and adversarial behaviors, by focusing on: (a) robust dyadic (i.e. pair-wise) interaction laws between particles (modeling UAV/UGV/UUV interactions) to achieve a rich collection of spatio-temporal patterns of cooperative behavior; (b) robust scalable dyadic pursuit laws as building blocks of adversarial interactions; (c) optimality principles underlying interactions laws that gives rise to cooperative and adversarial behavior; (d) insight from biological processes such as prey-capture behavior or pursuit, in echolocating bats, and in visual insects such as dragonflies;(e) verification of the theoretical advances in a laboratory test-bed consisting of commercial robotic platforms equipped with distributed agent-based control software, and an indoor GPS system based on ultrasonic ranging. Methods used in this research include techniques from geometric mechanics and control on Lie groups, and optimal control and game theory. The project has contributed basic results in pursuit laws as building blocks for cooperative and adversarial control. In collaboration with biologists, methods developed in this project are enhancing our understanding of prey-capture behavior in single and multiple echolocating bats.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 2010
- Accession Number
- ADA533583
Entities
People
- P.S.Krishnaprasad
Organizations
- University of Maryland