Objectives New technologies such as autonomous driving, intelligent networking, and V2X（vehicle-to-everything） are considered to be the key to driving changes in the future transportation system. Many studies have reached a consensus that human-driven vehicles, intelligent connected vehicles, and different levels of self-driving vehicles will coexist for a long period of time in the future. How to analyze the characteristics of the mixed traffic flow composed of these different types of vehicles is the key to rea‑lize the full promotion and application of vehicle-circuit coordination and autonomous driving technology. Simulation is an important method, but most of the existing microscopic traffic simulation tools do not have the ability to simulate the mixed traffic flow under the environment of vehicle-road cooperation.

Methods First, we build a vehicle driving behavior model under vehicle-road collaborative environment based on SUMO, an open-source micro-traffic simulation software, including three modules of driving control, driving advice and intelligent navigation, and six typical vehicle-road collaborative scenarios. Combined with the advantages of vehicle simulation software CARLA, which has high precision physical and sensor models, a collaborative co-simulation framework for expressway mixed traffic flow is proposed and named SC-V2XSim. The framework makes up for the defects that micro-traffic simulation does not have vehicle-road co-simulation function, can not access automatic driving algorithm and vehicle simulation does not have real traffic flow model. Second, the Shanghai-Hang-Ningbo expressway in Zhejiang Province is selected for testing and verification, and the characteristics of mixed traffic flow under different permeability conditions are simulated. Finally, a new generation information technology such as vehicle-road coordination is proposed to provide an analysis and evaluation method for the construction and development of the intelligent highway.

Results The experimental results show that holographic sensing, digital twin and V2X are three important factors for smart highway. On the other hand, with the increase of vehicles with vehicle-road collaboration function, the average road speed also gradually increases, and the average speed increases to 15% in the case of 100% V2X penetration.

Conclusions Compared with human-driven vehicles, intelligent connected vehicles have more reasonable driving behaviors. The proposed intelligent highway service level evaluation system, which considers vehicle-road coordination and other new generation information technologies, can provide a feasible analysis and evaluation method for the construction and development of smart highway.