Capacity Analysis and Cooperative Lane Changing for Connected and Automated Vehicles: Entropy-Based Assessment Method

Abstract

Connected and automated vehicles (CAVs) are poised to transform how we manage and control the existing traffic. CAVs can provide accurate distance sensing and adaptive cruise control which make shorter headway possible, and will eventually increase the roadway throughput or capacity. The vehicle-to-vehicle (V2V) communication technology equipment on CAVs allows vehicles to exchange information and form platoons more efficiently. This paper uses the intelligent driver model (IDM) as the behavior model to simulate CAVs in mixed traffic conditions with both CAVs and human-driven vehicles (HDVs) under different CAV penetration rates. A cooperative CAV lane-changing model is introduced to build more CAV platoons. The model develops two lane-changing algorithms. Partial CAV lane change (PAL) is applied at low CAV percentages, whereas full CAV lane change (FAL) is used at high CAV percentages. In addition, block entropy is employed as a performance measure for lane-changing results. The simulation experiments show that capacity will increase as the CAV percentage grows, and the peak growth rates occur in medium CAV percentage between 40% and 70%. The cooperative CAV lane-changing algorithm is found to decrease HDV–CAV conflicts remarkably by 37% as well as to marginally increase capacity by 2.5% under all CAV percentages. The simulation performance suggests that the threshold of CAV penetration rate for switching PAL to FAL is approximately 55%. Furthermore, it is demonstrated that block entropy can measure CAV lane-changing performance efficiently and represent capacity changes to some extent.