Research On Vehicle Interactive Behavior Characteristics And Models Based On Driving Style

With the improvement of motorization,the traffic safety problem is becoming more and more serious.Driving behavior state of drivers has a decisive influence on road traffic safety.The direct presentation of driving behavior in human-vehicle-road-environment system is vehicle following behavior and lane changing behavior,and the driver’s style is the supervisor and key factor affecting these two behaviors.It is very important to study the influence of driver’s style on vehicle driving behavior,explore the mechanism of vehicle-vehicle interaction,and establish a quantitative model of vehicle interaction behavior considering driving style for safe vehicle driving and smooth operation of traffic flow.The research content of this paper is divided into the following aspects.Firstly,the driving behavior data acquisition experiment was designed.A simulation driving experiment platform was built based on BMW3,12 drivers were recruited to collect the driving behavior data of the drivers in the simulation driving experiment.A self-reported driver behavior questionnaire was designed,and principal component analysis and K-means clustering method were used to complete the classification of driver styles.The standard of two behavioral data Windows was established,and the noise reduction was carried out on the collected data.Secondly,ANOVA method is used to analyze the data of drivers with different styles when following and changing lanes,and the variation rules of the main parameters of driving behavior characteristics are studied.It lays a foundation for the subsequent model establishment.Thirdly,vehicle following behavior characteristics are analyzed.According to the study on the main parameters of driving characteristics of drivers with different driving styles,a following behavior model considering driver styles is established.The model includes the expected distance model and acceleration model of vehicles with different driving styles in driving state.MATLAB numerical simulation software was used to conduct simulation experiments on the model to further explore the influence of following behavior on traffic flow.Finally,focusing on the new hybrid traffic flow in the intelligent networked traffic environment,the decision-making model of vehicle lane changing behavior is established.The dynamic risk model in the process of vehicle lane changing is introduced to establish the interaction relationship between autonomous vehicles and traditional vehicles in the hybrid traffic flow.Lane changing behavior of autonomous vehicles in mixed traffic flow is modeled based on game theory.If the cooperative lane changing behavior between autonomous vehicles is regarded as a non-cooperative game,the vehicles take their own driving state as the game income and seek the lane and running track with better driving conditions.The proposed lane change behavior model was simulated and verified by SUMO software.The simulation results show that the game lane change model has higher lane utilization and safety stability than the traditional gap threshold acceptance model.