Jam avoidance with autonomous systems

TL;DR

This paper compares autonomous and collective optimal velocity models for highway jam avoidance, finding that collective models with multiple interactions improve stability, but autonomous models with speed difference are often sufficient.

Contribution

It demonstrates that autonomous models with speed difference can achieve comparable stability to collective models with multiple interactions in jam avoidance.

Findings

Collective models with multiple interactions significantly increase stability.

Autonomous models with speed difference are nearly as effective as collective models.

Adding predecessors in collective models enhances stability.

Abstract

Many car-following models are developed for jam avoidance in highways. Two mechanisms are used to improve the stability: feedback control with autonomous models and increasing of the interaction within cooperative ones. In this paper, we compare the linear autonomous and collective optimal velocity (OV) models. We observe that the stability is significantly increased by adding predecessors in interaction with collective models. Yet autonomous and collective approaches are close when the speed difference term is taking into account. Within the linear OV models tested, the autonomous models including speed difference are sufficient to maximise the stability.