Adaptive Cruise Control with Safety Guarantees for Autonomous Vehicles

Magdici, Silvia; Althoff, Matthias · 2017 · OpenAlex-citations

DOI: 10.1016/j.ifacol.2017.08.418

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 challenge of designing an Adaptive Cruise Control (ACC) system for autonomous vehicles that simultaneously guarantees safety and ride comfort. While existing ACC systems improve traffic flow and comfort, they often fail to formally guarantee safety in emergency scenarios, such as when a leading vehicle suddenly brakes. The authors propose a novel control architecture that combines a nominal controller, based on Model Predictive Control (MPC), with a formally verified safety controller. This dual-layer approach ensures that the vehicle maintains a safe distance under all conditions while avoiding jerky maneuvers that compromise passenger comfort. The methodology involves a supervisory framework where the nominal MPC controller optimizes position tracking and minimizes jerk, assuming the leading vehicle maintains a constant velocity. However, since this assumption cannot guarantee safety against unexpected braking, the system continuously computes a required safe distance ($d_{safe}$) based on worst-case scenarios. If the nominal controller’s output fails to maintain this distance, the system engages a safety controller. The authors analyze four deceleration profiles for this safety maneuver: full deceleration, linear deceleration, exponential deceleration, and a proposed mixed deceleration profile. The mixed profile is selected because it balances the need for a short braking distance with the requirement for low jerk, ensuring comfort during emergency braking. The safe distance is computed analytically by determining the minimum distance between the host and leading vehicles over time intervals defined by their braking dynamics. The study evaluates the proposed framework against real traffic data. The results demonstrate that the system achieves good position and velocity tracking performance while strictly maintaining safety guarantees. The mixed deceleration profile allows for a gradual engagement of brakes, which significantly reduces jerk compared to constant full braking. Consequently, the system can handle sudden decelerations of the leading vehicle without causing uncomfortable maneuvers for passengers. The architecture ensures that control reverts to the nominal MPC as soon as the situation is no longer critical, leveraging the fact that most critical situations resolve quickly. The significance of this work lies in its ability to provide formal safety guarantees for autonomous vehicles without sacrificing comfort, a trade-off that plagues many existing ACC systems. By integrating a safety supervisor with a nominal MPC controller, the approach ensures collision avoidance even in the absence of inter-vehicle communication. The proposed mixed deceleration profile offers a practical solution for emergency braking that is both safe and comfortable. This framework contributes to the development of reliable autonomous driving systems by addressing the conflicting requirements of safety and comfort in a rigorous, mathematically verified manner.

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.

• following distance