AA/ECE/ME 597

Winter Quarter 2024

A networked dynamic system is a set of dynamical units that interact over an information-exchange network for its coordinated operation and behavior. Such systems have found many applications in diverse areas of science and engineering, including multiple space, air, land, and underwater vehicles, energy and power systems, physiology, and medicine. Currently, there is an active research effort underway in control and systems community to formalize these dynamical systems and lay out a foundation for their analysis and control synthesis. This course provides an overview of graph-theoretic techniques that have proven instrumental for studying networked dynamic systems. Specifically, we will look at the following topics: (1) network models (graphs, random graphs, random geometric graphs, state-dependent graphs, switching networks), (2) network properties (some useful combinatorial and algebraic properties of graphs), (3) dynamics over networks- theory and some applications, including agreement/consensus protocol and its various extensions, and (4) formation control. Other topics that will be covered- depending on the available time include: advanced algebraic methods in graph theory, biological networks, network controllability and observability, and games over networks. Suggested prerequisite is Linear Systems (AA/EE/ME 547) but if are taking linear systems concurrently and have solid background in linear algebra, you should be okay.

Class Time: (Wednesdays and Fridays) 1:00 pm - 2:20 pm

Class Room: Guggenheim Hall 204

Homework Assignments: Assigned every Friday (except the first week) and due the following Friday at 11:59pm on Class Canvas; homework grades will also be available on canvas

Homework Submission Link: Class Canvas

Project: There will be a project/paper/presentation for the course assigned during the second half of the course and due at the end of the quarter

Grading: Homework assignments will be 70% of the grade and 30% will be project

Office Hours: Thursdays 3:00-5:00 pm in AERB 130 (or instructor's office in Guggenheim Hall 311E)

You can access an online copy of the main textbook, referred as [MEBook2010] below here, provided by the UW library; other online resources or links to relevant books/papers will also noted.

Algorithms Illuminated, read about Euler's theorem history.

Course Schedule, Winter 2024

Date Topics Reference Notes & check these out
January 3 syllabus/logistics, graph theory basics, examples, algorithms, networkx Chapters 1 and 2 [MEBook2010] Algorithms Illuminated; A First Course in Network Science; Superspreaders of Covid-19; Controllability of brain network; your personal homework this week (no need to turn anything in) is to become familiar with python/juypter notebook/networkx or something similar, e.g., matlab/julia, to create, draw, and analyze networks/graphs, and be able to simulate dynamic systems described by differential equations
Jan 5 Euler's theorem, a few notions on graphs, connectivity, breadth-first-search Mehran Notes
Jan 10 No lecture due to MM travel will do a make up lecture/notes
Jan 12 node and edge connectivity, spectral graph theory, introduction to dynamics on adjacency and Laplacian, introduction to notions of control on networks Mehran Notes HW#1 assigned; see course canvas page
Jan 17 consensus protocol Chapter 3 [MEBook2010]; Mehran Notes Consensus protocol is a versatile distributed algorithm for achieving something that at the surface seems useless; but it is one of the most used module for devising a host of distributed algorithms for optimization, formation control, and estimation
Jan 19 consensus on directed networks Chapter 3 [MEBook2010]; Mehran Notes HW#2 assigned (see course canvas page); HW#1 due; check out: Naomi Leonard Podcast related to our discussion
Jan 24 Lyapunov analysis for networks Chapter 4 [MEBook2010]; Mehran Notes one of the original works in this area and a more recent work
Jan 26 More on Lyapunov analysis; unicyle formations Chapter 4 [MEBook2010]; Mehran Notes paper on planar collective motion by Sepulchre/Paley/Leonard
Jan 31 unicyle formations; formation specifications/graph rigidity Chapter 6 [MEBook2010]; Mehran Notes
Feb 2 formation control; Fax/Murray paper Chapter 6 [MEBook2010] HW#4 assigned; HW#3 due (both on canvas page)
Feb 7 games on networks/chip firing/more formation control Mehran Notes
Feb 9 state-dependent networks; network synthesis Mehran Notes HW#5 assigned; HW#4 due
Feb 14 network controllability
Feb 16 network controllability; network H2 performance and effective resistance HW#5 due
Feb 21 distributed least squares
Feb 23 distributed optimization no homework assigned this week
Feb 28 distributed estimation and filtering Mehran Notes
March 1 distributed estimation and filtering; Markov chains Mehran Notes HW#6 due
March 6 Project Presentations
March 8 Project Presentations HW#8 due; Project reports due on March 13