Negotiating Highway Interchange Traffic with a Decentralized Instability-Driven CBF-based Algorithm

TL;DR

The paper tackles high-density highway interchange lane swaps using a decentralized, instability-driven Predictor-Corrector CBF (PCCA) framework. It analyzes how tuning the inter-agent instability, via a steering-acceleration cost parameter, affects safety, responsiveness, and feasibility in a 2-DoF lane-change setting. Through elliptic CBFs and a QP-based safety filter, the approach demonstrates improved agility and robustness over virtual guard rails, with extensive Monte Carlo tests showing timely lane completion, manageable acceleration profiles, and resilience to non-responsive agents and V2V limitations. The work provides both theoretical feasibility results and practical insights for decentralized CBF-based coordination in mixed-traffic scenarios.

Abstract

In this paper we consider an interchange lane-swap scenario, a limited stretch of highway with two parallel lanes where most vehicles want to change lanes. We show that a particular decentralized Control Barrier Function based algorithm executes lane swaps efficiently, with minimal speed change, within the specified (short) road segment at high traffic densities (3,500 vehicles per hour per lane). Our main point is that controller tuning, the speed of inter-agent instability, plays a major role in the performance of the vehicle group. This is illustrated by comparing two different tunings of the controller and a third one where the lane swap is enforced by virtual guard rails. Like fighter jet dynamic instability improving maneuverability, the inter-agent instability improves agility of a group of vehicles. We emphasize that the controllers considered are decentralized: agents do not know if others want to change lanes or not.

Figures (7)

• Figure 1: The interchange zone between M-23 and I-94 highways in Michigan near Ann Arbor (source: Google Maps).
• Figure 2: Vehicles and elliptic CBFs with dimensions used in this paper.
• Figure 3: IDA-fast interchange lane swap: the non-transparent blue and red rectangles correspond to the same group of 6 vehicles at two time instants 2.3s apart. The "guardrail" lines are shown, but are not active for the IDA.
• Figure 4: IDA-fast run. Top plot: acceleration profiles of the 6 vehicles in focus. Bottom plot: their velocities in the $(x,y)$ plane.
• Figure 5: Virtual Guard Rails acceleration profiles for the same 6 vehicles and the same initial conditions used in Fig. \ref{['fig:acell']}