SE PhD Prospectus Defense of Yingqing Chen

TITLE: Towards Smarter Traffic: Advanced Control Solutions for Traffic Policy and Vehicle Management

ADVISOR: Christos Cassandras ECE, SE

COMMITTEE: Ioannis Paschalidis ECE, SE, BME Roberto Tron ME, SE Emiliano Dall’Anese ECE, SE

ABSTRACT: As urban environments become more congested and complex, the need for innovative traffic management solutions is critical. This research proposal explores advanced control strategies designed to enhance both the efficiency and safety of urban traffic systems, with a focus on two key areas.

The first part of the research tackles the challenge of optimizing traffic light control in grid-like urban networks. By developing adaptive, data-driven controllers, this study aims to improve vehicle flow across multiple intersections. Utilizing a stochastic hybrid system model combined with Infinitesimal Perturbation Analysis (IPA), the research introduces a novel framework for real-time traffic light optimization. This approach dynamically adjusts traffic light parameters in response to changing traffic conditions, accounting for transit delays, blocking effects, and diverse traffic patterns to ensure optimal network performance.

Building on the adaptive principles established in the first part, the second focus of this research extends these concepts to the domain of Connected and Automated Vehicles (CAVs). Specifically, it addresses the critical challenge of CAV navigation through single-lane roundabouts—key intersections within urban networks. By integrating Model Predictive Control (MPC) with Control Lyapunov-Barrier Functions (CLBFs), this work presents a control framework that not only minimizes travel time and energy consumption but also ensures safety through speed-dependent constraints. This approach resolves common issues in CAV control, such as infeasibility and myopic decision-making, and ensures smooth and safe navigation through interconnected zones.