View Uppdrag

Apply using URL

General

Headline

Organization/Company

Location

Assignment Type

Description

Keywords: Open Multi-Agent Systems, Signed Networks, Opinion Dynamics.

Supervisors: Pelin Sekercioglu (pelinse@kth.se), Martina Alutto (martina.alutto@polito.it)
Examiner: Angela Fontan (angfon@kth.se)

Context & Motivation: The study of synchronization in open multi-agent systems (OMAS) over signed graphs is motivated by the growing need to understand and control complex, dynamic networks that capture both cooperative and antagonistic interactions. Real-world systems such as social networks, robotic swarms, and information-exchange platforms are inherently open and evolve over time, with agents entering or leaving and relationships shifting between collaboration and conflict. Traditional consensus frameworks, while effective in static cooperative settings, fall short in explaining the richer dynamics arising in these evolving and antagonistic environments. In this project, we leverage a real data set to investigate synchronization in OMAS with signed interactions, thereby grounding the analysis in realistic and dynamic scenarios.

Background: OMAS are networks where agents and edges can dynamically be added to or removed from the system. They naturally arise in applications such as social networks, where the graph size changes with participant arrivals and departures. Many real-world scenarios involve antagonistic behaviors, such as social networks where agents compete or spread disinformation. A common framework for modeling both cooperation and antagonism in dynamical systems is that of signed graphs. 

In this project, we study the synchronization of OMAS over undirected signed graphs on a real data set. We consider systems where new nodes and edges can be added, and interconnections may switch between cooperation and antagonism.

The outline:
1. Literature review, including (signed) graph theory, consensus problems, switched systems. 
2. Demonstrate the proposed OMAS control algorithm in [1] using different network topologies (scale free, small world, Erdos-Renyi, tree, ...)
3. Demonstrate the proposed OMAS control algorithm on real-world datasets (candidate datasets are from https://networkrepository.com/ and https://snap.stanford.edu/data/).
4. Development of the algorithm from [1] in Python.
5. Experimental validation using the chosen data set.
6. (Optional) Extend the previous results by developing the proposed model of opinion dynamics, incorporating factors such as individual stubbornness or heterogeneous influence.

If the work progresses sufficiently, writing of a scientific article to submit to a leading control conference or journal.

## Your profile

Currently enrolled in a Master's program in electrical/control/robotics engineering or computer science, interest in working with real-world data, programming languages: Python.


## How to apply

Please send your CV and transcript of records via email to Pelin Sekercioglu (pelinse@kth.se) and Martina Alutto (martina.alutto@polito.it).

[1] P. Sekercioglu, A. Fontan, and D. V. Dimarogonas, “Stability of open multi-agent systems over dynamic signed graphs,” Accepted to Proc. IEEE Conf. on Dec. and Control 2025.

There are no records to display.

You don't have permissions to view these records.

Error completing request.

Deadline for application

Publish date

Credits

Application Channel

Application Email

Application URL

Application Documents

CV, transcript of records