Graph Connectivity for Multi-Agent Systems

Graph Connectivity for Multi-Agent Systems

Large-scale graphs

Graphs are a mathematical tool commonly used for studying multi-agent systems. One issue of importance is to quantify how “connected” a multi-agent system is, and this is done by studying the eigenvalues associated to special matrices describing the network interconnection. This project will explore how certain eigenvalues change with the addition and removal of connections in the network with the goal of finding optimal structures that balance the number of connections with given values of these eigenvalues.

Prerequisites:

  • control theory
  • Matlab and simulink
  • strong linear algebra a plus

Academic supervisor:

  • Prof. Daniel Zelazo mail

Duration: 1 or 2 semesters.