Empirical Study on the Acceleration/Deceleration Constraints Under Commercial Adaptive Cruise Control

This paper aims to identify the characteristics of acceleration/deceleration (accel/decel) constraints in com-mercial adaptive cruise control (ACC) and investigate their potential impact on traffic flow. We have collected empirical accel/decel data in Massachusetts and Georgia from various ACC-equipped vehicles including Tesla Model X/3, Civic, and Prius. Unlike constant values adopted in the literature so far, the observed maximum accel/decel rates across different ACC manufactures all suggest that the accel/decel constraints decrease linearly with vehicle speeds. Based on the data we also find: i) different ACC manufacturers seem to adopt different accel/decel constraints, ii) maximum accel rates do not vary significantly with the engine mode but differ in free-motion and car-following regimes iii) maximum decel rates show a strong correlation with minimum times to collision. We further demonstrate that accel/decel constraints in the ACC can contribute to speed overshootings/undershootings, causing capacity loss in the accelerating process and deteriorating congestion in the decelerating case. To reveal the underlying mechanisms, the paper develops an ODE model to analyti-cally approximate the speed overshootings/undershootings and correlate them with the accel/decel constraints as well as ACC control parameters. Besides speed over/undershootings, simulation results also show that accel/decel constraints in ACCs can undermine the string stability.