Login | Register

Failure Analysis in Multi-Agent Networks: A Graph-Theoretic Approach


Failure Analysis in Multi-Agent Networks: A Graph-Theoretic Approach

Rahimian, Mohammad Amin (2012) Failure Analysis in Multi-Agent Networks: A Graph-Theoretic Approach. Masters thesis, Concordia University.

[thumbnail of Rahimian_MASc_F2012.pdf]
Text (application/pdf)
Rahimian_MASc_F2012.pdf - Accepted Version


A multi-agent network system consists of a group of dynamic control agents which interact according to a given information flow structure. Such cooperative dynamics over a network may be strongly affected by the removal of network nodes and communication links, thus potentially compromising the functionality of the overall system. The chief purpose of this thesis is to explore and address the challenges of multi-agent cooperative control under various fault and failure scenarios by analyzing the network graph-topology. In the first part, the agents are assumed to evolve according to the linear agreement protocol. Link failures in the network are characterized based on the ability to distinguish the agent dynamics before and after failures. Sufficient topological conditions are provided, under which dynamics of a given agent is distinguishable for distinct digraphs. The second part of this thesis is concerned with the preservation of structural controllability for a multi-agent network under simultaneous link and agent failures. To this end, the previously studied concepts of link and agent controllability degrees are first exploited to provide quantitative measures for the contribution of a particular link or agent to the controllability of the overall network. Next, the case when both communication links and agents in the network can fail simultaneously is considered, and graphical conditions for preservation of controllability are investigated.

Divisions:Concordia University > Gina Cody School of Engineering and Computer Science > Electrical and Computer Engineering
Item Type:Thesis (Masters)
Authors:Rahimian, Mohammad Amin
Institution:Concordia University
Degree Name:M.A. Sc.
Program:Electrical and Computer Engineering
Date:30 May 2012
Thesis Supervisor(s):Aghdam, Amir G.
Keywords:Multi-Agent Networks, Controllability, Graph Theory, Algebraic Graph Theory, Linear Systems, Networked Dynamics, Agreement Dynamics
ID Code:974929
Deposited On:06 Nov 2012 20:42
Last Modified:18 Jan 2018 17:39
All items in Spectrum are protected by copyright, with all rights reserved. The use of items is governed by Spectrum's terms of access.

Repository Staff Only: item control page

Downloads per month over past year

Research related to the current document (at the CORE website)
- Research related to the current document (at the CORE website)
Back to top Back to top