Connected Vehicle Data-Aided Ramp Metering for Distant Bottlenecks Induced by Traffic Accidents

Abstract

This paper explores the enhancement of ramp metering strategies using data from connected vehicles (CVs), specifically targeting non-recurrent bottlenecks caused by traffic accidents. Traditional ramp metering, which relies on fixed sensors for feedback control, is often ineffective in managing these unpredictable, distant bottlenecks. However, this issue may become more tractable in a CV environment, as CVs can serve as mobile sensors, providing broader spatial coverage and generating more granular speed and position data. In this paper, we proposed using CV data to estimate bottleneck-related traffic states for implementing feed-forward ramp metering algorithms that addresses congestion around distant bottlenecks. Our method was tested and validated in a micro-simulation model, across various CV scenarios, with different market penetration scenarios. The findings indicate improved traffic mobility and more effective responses to accident-induced bottlenecks in a CV-enhanced ramp metering system both at the local and system level.