Proactive Longitudinal Control to Assist Lane Changes of Human-Driven Vehicles in Mixed Traffic: Human-Emulation Approach

Liu, Yongyang; Zhou, Anye; Wang, Yu; Peeta, Srinivas · 2026 · Crossref

DOI: 10.1007/s42154-025-00357-9

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 maintaining traffic smoothness and safety in mixed traffic environments where connected and autonomous vehicles (CAVs) coexist with human-driven vehicles (HDVs). Specifically, it targets the disruption caused by HDV lane changes near CAV platoons, which often induce traffic oscillations and reduce platoon performance. The authors propose a Human-Emulation-based Proactive Longitudinal Control (HPC) strategy designed to assist HDVs in executing lane changes smoothly. By emulating the cooperative yielding behavior typical of human drivers, the HPC aims to make CAV intentions legible to HDV drivers, thereby reducing mental workload and enhancing safety. The HPC framework operates in three phases and utilizes two primary control models. First, a Transformer-based behavior predictor analyzes kinematic data (position, speed, acceleration) of nearby vehicles to anticipate HDV lane changes. If a lane change is predicted, the system activates a lane-change assistance model formulated as a Markov Decision Process (MDP) and solved using Proximal Policy Optimization (PPO). This model encourages the CAV to decelerate proactively, creating a safe gap and signaling yielding intent. If no lane change is predicted, the CAV employs a multi-anticipative car-following model, specifically the Extended Intelligent Driver Model (EIDM), to maintain cooperative platooning. The reward function for the PPO algorithm prioritizes traffic smoothness, motion legibility, speed convergence, and collision avoidance. The effectiveness of the HPC strategy was evaluated through driving simulator experiments with 36 participants and numerical simulations. The simulator results demonstrated that the HPC strategy significantly improved traffic smoothness, achieving an 84.3% reduction in speed perturbation compared to baseline strategies. It also enhanced safety for HDV drivers, resulting in a 32.6% increase in time-to-collision, and improved mental comfort, evidenced by a 13.9% decrease in self-reported workload. Numerical experiments further confirmed that the HPC reduces speed perturbation by 19.1% across various scenarios, driver types, and control setups. The study concludes that incorporating human factors into CAV control systems through legible, human-emulated motions improves interactions in mixed traffic. The HPC strategy not only mitigates traffic oscillations but also facilitates safer and smoother lane changes for HDVs. A significant finding is that HDVs exhibit a propensity for additional lane changes in front of CAVs controlled by HPC, potentially exploiting the CAV’s cautious and legible behavior. This highlights the need for CAV systems to balance caution with assertiveness to prevent "bullying" behaviors while maintaining overall traffic efficiency and safety.

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 2026-06-09; verification: verified.

Topics

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

• teleoperation remote driving

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Theoretical Contribution: computational model