Followerstopper Revisited: Phase-space Lagrangian Controller for Traffic Decongestion

摘要

This paper revisits Followerstopper, a phase-space-based control system that had demonstrated its ability to mitigate emergent traffic jams due to stop-and-go traffic during rush hour in the mixed-autonomy setting. Followerstopper was deployed on an autonomous vehicle. The controller attenuates the emanant traffic waves by regulating its velocity according to the relative distance and velocity of the leader car. While regulating the velocity, the controller also prevents the collision of the ego vehicle with the lead vehicle within the range specified by the controller's design parameter. The controller design is based on a configurable quadratic curve on relative distance-relative velocity phase-space that allows the transition of the regulated velocity from (i) no modification of input, (ii) decelerating to match the leader's velocity (iii) braking to avoid any imminent collision. In this paper, we explore the phase-space properties of Followerstopper and provide a detailed description of a nonlinear control law that regulates the reference input to Followerstopper within the physics-informed boundaries. We also provide a new discussion on the nominal control law that regulates the reference speed to Followerstopper to avoid unrealistic and unsafe acceleration.