Eco-driving control for CAVs at signalized intersections: Adapting to traffic uncertainties

• Published in: Transportation research. Part D, Transport and environment Transport and environment, 2024-07, Vol.132, p.104270, Article 104270
• Main Authors: Lu, Kaiming, Chen, Yanyan, Tong, Yao, Zhang, Jian, Luo, Ying, Wang, Jiachen
• Format: Article
• Language: English
• Subjects: Connected and automated vehicles; Eco-driving control; Signalized intersections; Traffic uncertainties
• ISSN: 1361-9209; 1879-2340
• DOI: 10.1016/j.trd.2024.104270

Eco-driving control for connected and automated vehicles (CAVs) presents significant energy-saving potential by enabling smooth, non-stop travel through signalized intersections. This study addresses the challenge of traffic uncertainties under congestion conditions, which often lead to trajectory planning errors and increased energy consumption. We propose a hierarchical eco-driving approach for CAVs that incorporates both upper-layer trajectory planning and lower-layer velocity control. The upper layer employs a dynamic programming-based model for time-based optimal eco-driving reference trajectory planning, focusing on objectives of energy efficiency and high mobility. In the lower layer, a model predictive control (MPC)-based two-mode velocity controller is formulated to adeptly handle infeasible or unreasonable reference trajectories. Evaluation results demonstrate that the proposed approach outperforms existing strategies, achieving average improvements of 3.99% in CO2 emissions, 3.65% in fuel consumption, and 2.61% in vehicle mobility. We suggest that a comprehensive evaluation framework for eco-driving performance of CAVs should be established with respect to diverse traffic conditions, regulating CAV market entry from an environmental perspective.