Title: Distributed Optimization Architectures for Large-Scale Decision-Making

Date: Friday, July 18th, 2025

Time: 1:00 - 3:00 pm EST

Location: Coda C1215 Midtown

Meeting Link: https://gatech.zoom.us/j/91891408754

Augustinos Saravanos

Machine Learning PhD Student

School of Electrical and Computer Engineering
Georgia Institute of Technology

Committee

1. Dr. Evangelos Theodorou (School of Aerospace Engineering, Georgia Tech; Advisor)

2. Dr. Arkadi Nemirovski (School of Industrial and Systems Engineering, Georgia Tech)

3. Dr. Yao Xie (School of Industrial and Systems Engineering, Georgia Tech)

4. Dr. Justin Romberg (School of Electrical and Computer Engineering, Georgia Tech)

5. Dr. Efstathios Bakolas (Department of Aerospace Engineering and Engineering Mechanics, UT Austin)

Abstract

As the scale and complexity of modern multi-agent systems are rapidly increasing, the development of scalable and reliable decision-making algorithms is becoming increasingly vital. This thesis introduces a series of novel distributed optimization, control, and learning-based architectures that address these challenges, ensuring computational efficiency, scalability, robustness under uncertainty and interpretability. The main contributions can be summarized as follows: (i) Two novel distributed dynamic optimization architectures for large-scale multi-agent control are introduced, providing state-of-the-art scalability for optimal control in autonomous systems and demonstrating their effectiveness through hardware experiments. (ii) A family of scalable decentralized methods for distribution steering in multi-agent stochastic systems is presented, providing trade-offs between safety capabilities and computational efficiency. (iii) A model predictive control version of the decentralized distribution steering framework is then proposed, enabling its application to real-world multi-agent systems operating under uncertainty. (iv) A hierarchical distribution optimization architecture is presented for very-large-scale clustered multi-agent systems, which exploits underlying hierarchies to achieve improved scalability and robustness. (v) Finally, a deep learning-aided distributed optimization architecture for large-scale quadratic programming is introduced, demonstrating substantially improved performance over traditional optimizers and strong generalization to large-scale problems.