How to Estimate the Traffic Propagation under the Influence of an Administered Autonomous Vehicle?

Abstract: Traffic State Estimation (TSE) involves estimating the propagation of traffic characteristics along a road section based on specific traffic observations. The use of Connected and Autonomous Vehicles (CAVs), particularly as Administered Autonomous Vehicles (AAVs), presents a promising approach to influencing traffic state propagation. This has the potential to improve traffic efficiency, reduce fuel consumption, enhance stability, and optimize control in time-space regions. However, methods for estimating traffic states under the influence of AAVs remain underdeveloped. This study introduces an online traffic physics-based traffic state estimation method that accounts for the presence of an AAV. First, a digital traffic state propagation representation method is proposed, aligning with the developed estimation algorithm developed in this paper. A Recurrent Earliest Intersection (REI) algorithm is then formulated to estimate traffic state propagation within a trapezoidal time-space region. A double coordinate system is incorporated to capture the impact of an AAV on traffic state propagation. To accurately assess the traffic state around the AAV, interactions between the AAV and surrounding traffic are analyzed, providing inputs to the developed TSE algorithm in this paper. Finally, a comprehensive Traffic State Propagation Estimation method under the influence of an AAV (TSPE_aav) is developed by coordinating dual time-space regions, considering AAV-traffic interactions. Numerical experiments demonstrate the effectiveness of TSPE_aav across various traffic scenarios, including freeway and arterial networks, and illustrate the potential effect of AAVs in traffic control.