Study On The Influence Of Driving Experience On Car-Following Behavior

Car-following behavior is an important part of modern traffic flow theory,which has great significance in analysing and researching on road capacity,traffic safety control,road evaluation management and so on.Then the drivers as the judge and decision makers in the process of dealing with traffic information,their experience influences the operation of the whole operating system,which is the most active factor in the traffic operation.Exploring the impact of driving experience on car-following behavior played an important role in traffic flow theory’s complement and improvement.First of all,this paper introduces various test methods about collecting the data of car-following in the past traffic flow survey,and introduced the analysis system of car-following behavior based on vehicle-mounted lidar equipment.Then carried out a series of field car-following tests for drivers with different experiences.After the validity of the test data,a large number of car-following behavior data that reflecting the characteristics of the driver’s personal experience are obtained.Secondly,the effects of driving experience on stable car-following behavior and acceleration/deceleration car-following behavior are studied respectively.The results of the stable car-following state study show that the inexperienced drivers’ car following distance is generally longer than experienced drivers,the correlation coefficient between the driver’s accumulated driving mileage and the distance from the car is negative,and the distance between inexperienced drivers and experienced drivers was respectively obey ln N(2.985,0.828)and ln N(2.874,0.954)lognormal distribution;inexperienced drivers of the headway is generally higher than experienced driver too,the correlation coefficient between driver’s accumulated driving mileage and the headway is also negative,and the headway between inexperienced drivers and experienced drivers was respectively obey N(6.092,0.349)and N(5.900,0.250)lognormal distribution.The results of the acceleration/deceleration car-following state study show that the inexperienced drivers’ reaction time is generally longer than experienced drivers,the correlation coefficient between the driver’s accumulated driving mileage and the reaction time is negative,and the reaction time between inexperienced drivers and experienced drivers was respectively obey N(2.940,1.011)and N(2.280,0.993)normal distribution under acceleration,while under deceleration respectively obey Γ(3.810,1.506)and Γ(1.831,1.522)gamma distribution.Under two different car-following states,the reaction time has logarithmic curve relation with vehicle speed and headway,and decreases with the increase of speed,and increases with the increase of vehicle headway.Finally,the safety distance model is selected as the evaluation model,based on the analysis of driver reaction time under different conditions,the parameters of the model are calibrated by nonlinear regression analysis,and the fitting degree of the model is obtained.Then the measured headway value is compared with the calculated value of the model.Results show that after calibration,the model can well reflect the corresponding vehicle under different driving experience with car-following situation,to consider the driver experience with car behavior research provides the theoretical basis and model reference.