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

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.