Research On Lane Changing Decision Method Of Connected Vehicles Based On Game Theory

In the process of driving,drivers take different driving behaviors according to the changes of road environment and surrounding vehicle information.In the two driving behaviors of lane changing and following,because of the complexity of the lane changing process,it is more likely to cause traffic accidents due to the wrong decision of lane changing.In the connected environment,the driving assistance system can provide decision-making assistance for the driver's lane changing behavior through the comprehensive perception of the surrounding environment and vehicles.Most of the lane change decision-making models adopted by the existing lane change driving assistance system are established based on the kinematics rules of vehicles.They can only give early warning of lane-changing danger before drivers change lane and fail to consider the characteristics of drivers,which makes it difficult to reflect the changes of drivers' driving behaviors in the process of lane change.In view of the above problems,this paper proposes a lane change driving behavior decision-making model based on game theory,which is mainly based on the idea of game theory and considering the driving styles of different drivers at the same time,and conducts experimental verification on its performance.Based on the in-depth study of the game theory and the demand for lane changing behavior,the pure strategy game model of simple lane changing decision in typical lane changing scenario is established by using the static non-cooperative game method under complete information.After judging the connected vehicle having lane-changing motivation or not,we made the pure strategy model of lane changing as the security pay-off according to the characteristics of lane changing behavior,proposed the calculation method of game benefit function considering the time benefit,and the feasibility of the benefit function was verified by Matlab experimental simulation model.The style coefficient is introduced into the participants' strategic return,we gave the different weights to the decision return of different driving styles,and analyzed and calculated the game return matrix,then we solved the decision model.After using the SUMO simulation software combined with the lane-changing data from processing of the NGSIM data set,we simulated the vehicle automated driving permeability in different environments to compare the contrast of the proposed model with CA model through the traffic,lane changing times,accidents,and such as the results of data analysis,then we verified the validity and stability of lane changing decision method in this paper.Simulation results show that,in the traditional,hybrid,and unmanned environment,this paper proposed a lane changing decision model based on game theory,which can improve time payoff on the premise of safety and convert safety margin into time payoff.The validity and stability of the model is good.Compared with the real data,the suggested accuracy of the model can reach 87.6%.The research of this paper shows that the driving assistance method for lane changing decision-making proposed in this paper can provide driving suggestions for lane changing vehicles in the course of lane changing,assist vehicles to complete lane changing and ensure the safety and stability of lane changing,and it provides a feasible decision-making method for the future establishment of the driving assistance system for lane changing process.