Evaluating Model Mismatch Impacting CACC Controllers in Mixed Traffic Using a Driving Simulator
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses the critical challenge of model mismatch in Cooperative Adaptive Cruise Control (CACC) systems operating in mixed traffic environments, where automated vehicles share the road with manually driven vehicles (MDVs). As automated vehicle penetration increases, CACC controllers must predict the behavior of legacy vehicles to ensure safety and comfort. However, discrepancies between the controller’s assumed models of human behavior and actual driver actions can lead to inefficient control inputs, increased discomfort, and potential collisions. The authors propose a framework to evaluate these impacts by interfacing a centralized controller with a driving simulator, allowing for the inclusion of real human drivers in the experimental loop. The study employs a centralized Model Predictive Control (MPC) algorithm designed for a braking scenario where vehicles approach an obstacle. The controller predicts MDV behavior using a simplified model that accounts for perception response times and braking dynamics. To test the system, the authors used a simulation framework integrating MATLAB for the controller and the Swedish National Road and Transport Research Institute’s (VTI) driving simulator for human input. Two cases were evaluated: Case A, where the actual MDV behavior was simulated using a modified Intelligent Driver Model (IDM), and Case B, where actual braking inputs were provided by six human participants using a steering wheel and pedals. The controller attempted to coordinate braking for the CACC vehicle to avoid collisions while minimizing jerk and acceleration deviations. The results demonstrate that model mismatch significantly degrades system performance. In Case A, where the assumed model matched the simulated IDM behavior, the controller achieved a 100% collision avoidance rate with a discomfort metric of 6.66. In contrast, Case B, involving real human drivers, resulted in a collision avoidance rate of only 53.57% and a discomfort metric nearly doubling to 12.76. The authors attribute this failure to discrepancies in perception response times, braking capacity, and jerk sustainability between the assumed model and actual human behavior. Additionally, computational delays were observed; the optimization process took up to two seconds, far exceeding the 0.1-second update rate, forcing the system to rely on outdated control inputs from a buffer, which further contributed to collision risks. The significance of this work lies in its validation of the necessity for robust control strategies that account for human variability in mixed traffic. The findings indicate that model mismatch is non-negligible and can lead to severe safety and comfort issues. The proposed framework offers a unique method for evaluating CACC controllers by incorporating real human drivers, highlighting the limitations of purely simulated or theoretical assessments. The study concludes that future CACC designs must address computational efficiency and incorporate more robust or adaptive models to handle the unpredictability of human driving behavior effectively.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-06-20 |
| archive | success | semantic_scholar | — | — | 6 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| enrich | success | openalex | — | — | 1 | 2026-06-20 |
| promote | success | — | — | — | 1 | 2026-06-20 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-20 |
| verify | success | — | — | — | 1 | 2026-06-26 |
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.
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