By analyzing the influence of time and safety factors on the behavior strategies of vehicles on the road, a network game evolution model between drivers that considers the behavior strategies of the driving vehicle itself and its neighbors is constructed, and the competition relationship between different types of cars is studied. The influence of the proportion of driving vehicle types on the potential risk of the road is also discussed. This paper presents a guidance algorithm for vehicle dynamic behavior preference information. The correctness of the algorithm is verified by an example. Research shows: The choice of behavior strategies, such as speeding and lane changing, is related to the expected benefits of time, safety, and neighboring vehicle strategies, and the critical value of payable benefits is obtained. The higher the proportion of aggressive vehicles on the road, the greater the potential risk on the road. Whether there is a vehicle in the adjacent lane of the driving vehicle will affect the type of driving vehicle. Information guidance helps to stabilize the state of vehicles on the road, and the policy transition probability also helps stabilize the form of vehicles cars on the road. Still, information guidance has a more significant impact on the transition of vehicle types. Finally, the guidance strategy of managers is given when the road is smooth and congested.