Intersection passing strategies for human-driven and autonomous vehicles in mixed traffic using DEA.

Shen, J; Zhou, Z; Fu, G; Wang, Y · 2025 · PubMed Central

DOI: 10.1371/journal.pone.0321566

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of optimizing intersection passing strategies in mixed traffic environments containing both human-driven vehicles (HVs) and connected autonomous vehicles (CAVs). As CAVs are gradually introduced into transportation networks, the transition period will feature mixed traffic where HV unpredictability limits CAV efficiency gains. Traditional control methods, such as actuated signals or First-Come-First-Served (FCFS) policies, often fail to comprehensively optimize intersection efficiency because they rely on limited factors. To resolve this, the authors propose a right-of-way optimization model using Data Envelopment Analysis (DEA) to determine the optimal traffic flow sequence by balancing multiple objectives. The methodology employs a multi-objective DEA evaluation framework. The model considers six specific influencing factors: average speed, number of cars, CAV penetration rate, queuing patterns (categorized by vehicle arrangement), left-turn rate, and number of buses. The optimization targets three primary metrics: per capita delay, travel time, and traffic volume. Per capita delay is prioritized to account for high-capacity buses, while travel time and traffic volume ensure overall throughput and timeliness. The study utilizes the Crossing Efficiency Evaluation Method (CREE) to calculate cross-benefits by interchanging weight evaluations, thereby determining an objective passing order for each traffic stream. Vehicles are grouped into platoons based on headway, and the roadside unit adjusts signal phases to release these platoons in the calculated optimal order. The proposed strategy was validated using SUMO simulation software and compared against existing benchmarks, including actuated control and FCFS strategies. The simulation results demonstrate that the DEA-based optimization strategy significantly shortens both per capita delay and travel time at intersections. By comprehensively weighing multiple traffic characteristics rather than relying on single indicators, the model improves the overall efficiency of traffic flow compared to conventional methods. The significance of this research lies in providing a holistic approach to intersection management during the transitional phase of autonomous driving. Unlike previous studies that focus primarily on CAV-specific optimizations or require extensive data training for reinforcement learning, this method offers a rapid and accurate determination of passing orders without large-scale training. It effectively balances the needs of different vehicle types, particularly prioritizing public transport through per capita delay metrics, thereby enhancing the operational efficiency of urban road networks in mixed-traffic scenarios.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• traffic density