Evaluate the Connected Autonomous Vehicles Infrastructure using Digital Twin Model Based on Cyber-Physical Combination of Intelligent Network

Abstract

With the increment of connected vehicles, the level of intelligence becomes more and more irregular, so the difficulties of determining the dynamic safety of self-driving in mixed-transport flow have increased significantly. To solve the problems such as reliability, human-car-road perception, decision making, and control coordination assessment in an intelligent networked environment, this article established a multi-source dynamic game model to carry out the measurement of autonomous vehicle dynamics model, control estimation, decision strategy, forward and backward safety mechanism, and planning of mixed-traffic flow route. The digital twin has real-time, synchronous evolution, and interactivity with a semi-physical environment and a hardware-in-the-loop (HIL) model to control the accuracy of dynamic safety decisions for smart connected vehicles. This all process developed by combining with vehicle-to-everything (as a physical entity) and smart simulation test technology (as a virtual entity), which understands the compound and dynamic safety decision objects such as multi-agent view, multi-source data communication, vehicle switching, V2V transmission, and V2R synchronization for connected autonomous vehicles (CAVs) in the mixed-traffic flow atmospheres.