Dissertation
Topological characterizations of the controllability of complex networks
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Spring 2020
DOI: 10.17077/etd.005338
Abstract
Complex networks composed of units dynamically interacting among themselves have been found broad applications in brain networks, social networks, multi-agent networks, and power networks. Owing to tremendous application potential, growing research has been devoted to investigating structural and functional properties of complex networks. The focus of this dissertation is to characterize classical controllability and generalized controllability of complex networks from graph-theoretic perspectives.
Classical controllability of complex networks is studied in Chapter 2. By exploring connections among leaders and followers, leader selection algorithm is developed through graph topology to guarantee networked systems with both cooperative and antagonistic interactions to be controllable. Energy-related controllability is investigated from Chapter 3 to Chapter 5. In chapter 3, energy-related controllability of complex networks is studied through individual performance of leaders. Followed by the results in Chapter 3, a novel leader selection method using graph distance to ensure network controllability with better energy-related performance is proposed in Chapter 4. Further, Chapter 5 analyzes energy-related performance of composite networks via graph product. Chapter 6 introduces a relaxed controllability referred to as herdability. Generalized Equitable partition is utilized to characterize herdability of complex networks in Chapter 6, while graph walk is applied to leader selection to ensure herdability of complex networks in Chapter 7. Chapter 8 characterizes structural balance of signed networks via algebraic topology. Chapter 9 concludes the dissertation by summarizing the work and discussing the open problems that require further investigation.
Details
- Title: Subtitle
- Topological characterizations of the controllability of complex networks
- Creators
- Baike She
- Contributors
- Zhen Kan (Advisor)Shaoping Xiao (Advisor)Er-Wei Bai (Committee Member)Siddhartha Mehta (Committee Member)Venanzio Cichella (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Mechanical Engineering
- Date degree season
- Spring 2020
- DOI
- 10.17077/etd.005338
- Publisher
- University of Iowa
- Number of pages
- xi, 141 pages
- Copyright
- Copyright 2020 Baike She
- Language
- English
- Description illustrations
- color illustrations
- Description bibliographic
- Includes bibliographical references (pages 133-141).
- Public Abstract (ETD)
Complex networks composed of units dynamically interacting among themselves have been found broad applications in brain networks, social networks, multi-agent networks, and power networks. Owing to tremendous application potential, growing research has been devoted to investigating structural and functional properties of complex networks. The complexity of complex networks lays down a challenge for analysis and design networked dynamical systems through traditional tools. Hence, the focus of this dissertation is to characterize properties and performance of networked dynamical systems from graph-theoretic perspectives. The gaps between system performance and network structures are built to facilitate the design of networked systems, which can be applied to social network analysis and swarm robotics.
- Academic Unit
- Mechanical Engineering
- Record Identifier
- 9983949498002771
Metrics
119 File views/ downloads
75 Record Views