CAVSim: A Microscope Traffic Simulator for Connected and Automated Vehicles Environment

Abstract

Connected and automated vehicles (CAVs) are expected to play a vital role in the next-generation intelligent transportation system. In recent years, researchers have proposed various cooperative driving methods for CAVs' decisions and planning, and there is an urgent need for a suitable and unified traffic simulator to evaluate and test these methods. However, existing traffic simulators have three critical deficiencies for CAV simulation needs: (1) most of them are from the perspective of traffic flow simulation and have strong simplification and assumptions for vehicle modeling; (2) CAVs are different from traditional human-driven vehicles (HVs) and have new properties, which require new driving models; (3) the existing traffic simulators are inconvenient to deploy the emerging cooperative driving methods because their modeling of the traffic system architecture is traditional. In this paper, we introduce CAVSim, a novel microscope traffic simulator for the CAV environment, to address these deficiencies. CAVSim is modularly developed according to the emerging architecture for the CAV environment, emphasizes more detailed driving behaviors of CAVs, and highlights the decision and planning components in the CAV environment. With CAVSim, researchers can quickly deploy decision and planning methods at different levels, evaluate and test their performance, and explore their impact on the traffic flow in the CAV environment.