Characterizing the driving behavior of manual vehicles following autonomous vehicles and its impact on mixed traffic performance

Abstract

An important issue for mixed traffic conditions, in which autonomous vehicles (AVs) and manual vehicles (MVs) coexist, is to analyze various vehicle interactions caused by different driving behaviors. Understanding the responsive behavioral characteristics of the following MV affected by the maneuver of the leading AV is a backbone in evaluating mixed traffic performance. The purpose of this study is to characterize the driving behavior of MVs following AVs in mixed-traffic situations. To characterize vehicle interactions between AVs and MVs, this study conducts multi-agent driving simulation (MADS) experiments, which can synchronize the space and time domains on the road by connecting two driving simulators. A maneuvering control logic for AV driving, which is used for MADS, is developed in this study. The driving behavioral data of MVs following AVs obtained from MADS are used to modify the parameters associated with the intelligent driver model (IDM). The IDM is a microscopic car-following model to represent the longitudinal following behavior of vehicles. This study identifies how the MV following AV would be different from the case where the MV follows MV. The results show that the average time headway of the following MVs in the AV-MV pair increased by 13.9% compared to the MV-MV pair. However, the maximum acceleration and average deceleration decreased by 44.45% and 4.89%, respectively. The proposed IDM for MV following AV was further plugged into a microscopic traffic simulation platform. VISSIM simulations were conducted to identify the difference in driving behavior between the proposed IDM and the original IDM. The outcome of this study is expected to simulate the maneuvering behavior of MV more realistically in the mixed traffic stream.