EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

1866. Analytical Strategy for Optimizing Mixed Traffic Networks with Semi-Autonomous Vehicles

Invited abstract in session MC-56: Traffic flow modeling , stream Transportation.

Monday, 12:30-14:00
Room: S04 (building: 101)

Authors (first author is the speaker)

1. Hao Guan
CEE, National University of Singapore
2. Xiangdong Chen
Civil and Environment Engineering, National University of Singapore
3. Qiang Meng
Civil and Environment Engineering, National University of Singapore

Abstract

Despite recent progress in semi-autonomous vehicle (semi-AV) development, achieving an optimal autonomous traffic system faces challenges due to the low penetration rate of semi-AVs and their need for human oversight. To improve traffic flow in such environments with human-driven vehicles (HVs) and semi-AVs, we advocate the use of dedicated lanes and develop an innovative framework to optimize the network performance by planning the mode choice of semi-AVs (either human-driven or auto-driven). Specifically, we introduce a three-dimensional Macroscopic Fundamental Diagram (3D-MFD) to characterize the complex flow dynamics in mixed traffic. Our analysis identifies that an optimal proportion of human-driven and auto-driven vehicles can significantly maximize network flow. Inspired by this key finding, we construct an optimization model to regulate semi-AVs’ mode choice, aiming to optimize vehicle proportions within mixed traffic networks. Although solving the model presents challenges due to its nonlinearity and integer constraints, especially in large networks with numerous intersections, the unique structure of the proposed 3D-MFD enables us to derive an analytical solution. This solution significantly reduces the computational complexities of the optimization model, allowing the resulting control strategy to be implemented in real time and substantially improves network performance.

Keywords

Status: accepted

Back to the list of papers