Contraction and partial contraction : a study of synchronization in nonlinear networks
Name
61134167-MIT.pdf
Description
Full printable version
Size
6.38 MB
Format
Adobe PDF
Checksum (MD5)
3942db7a31c14d3e4bb730a65638a6a8
Author(s)
Wang, Wei, 1972 Oct. 17-
Advisor(s)
Jean-Jacques E. Slotine.
Date Issued
2005
Publisher
Massachusetts Institute of Technology
Abstract
This thesis focuses on the study of collective dynamic behaviors, especially the spontaneous synchronization behavior, of nonlinear networked systems. We derives a body of new results, based on contraction and partial contraction analysis. Contraction is a property regarding the convergence between two arbitrary system trajectories. A nonlinear dynamic system is called contracting if initial conditions or temporary disturbances are forgotten exponentially fast. Partial contraction, introduced in this thesis, is a straightforward but more general application of contraction. It extends contraction analysis to include convergence to behaviors or to specific properties (such as equality of state components, or convergence to a manifold). Contraction and partial contraction provide powerful analysis tools to investigate the stability of large-scale complex systems. For diffusively coupled nonlinear systems, for instance, a general synchronization condition can be derived which connects synchronization rate to net- work structure explicitly. The results are applied to construct flocking or schooling models by extending to coupled networks with switching topology. We further study the networked systems with different kinds of group leaders, one specifying global orientation (power leader), another holding target dynamics (knowledge leader). In a knowledge-based leader-followers network, the followers obtain dynamics information from the leader through adaptive learning. We also study distributed networks with non-negligible time-delays by using simplified wave variables and other contraction-oriented analysis. Conditions for contraction to be preserved regardless of the explicit values of the time-delays are derived.
(cont.) Synchronization behavior is shown to be robust if the protocol is linear. Finally, we study the construction of spike-based neural network models, and the development of simple mechanisms for fast inhibition and de-synchronization.
Description
Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 2005.
Includes bibliographical references (p. 121-128).
Subjects
Mechanical Engineering.
MIT Department
Massachusetts Institute of Technology. Department of Mechanical Engineering
Terms of Use
M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission.
Persistent DSpace Link
