Research On Car Following Model Considering Preceding Vehicle's Lane-Changing Process

Car following and lane changing are two common driving behaviors in the traffic flow.Preceding vehicle's lane changing is affected by the surroundings,and will have greatly influence the followers' driving decision.The existing car-following theory doesn't fully take into consideration thar the followers' driving behavior may change during lane-changing process,and the related research on lane changing is rarely meticulously discussed the influence the on the followers at a micro level.Therefore,the car-following model considering the lane-changing process can improve the existing models' adaptability and truly reflect the driving behavior in comlex traffic flow.Based on the NGSIM(Next Generation Simulation)dataset,the influence on the following vehicles of the lane-changing vehicle on the target lane and the current lane is analyzed,and on the basis of the FVDM,the car-following model during lane-merging(LM)process and lane-passing(LP)process are proposed respectly.The stability analysis on the models are operated and the numerical simulation are conducted to study the models' rationality.Finally the NGSIM lane-changing trajectory data is used to evaluate the models' validity.(1)Based on the driving state data of the surrounding vehicles,which are extracted from NGSIM dataset,the hypothesis test statistical method is used to analyze the impact of the lane-changing vehicle on the following vehicles on the target lane and the current lane respectively,adopted the average velocity,the velocity's standard deviation and the velocity'variation coefficient.(2)The LM-FVDM is proposed.Three lane-changing stimulus for the following vehicle on the target lane are difined when the vehicle is lane-merging into the target lane,and the competitive level function is introducted with the consideration of drivers' competition coefficient of the follower and lateral separations of the lane-changing vehicle.Based on the principle of FVDM model,the model of driver car-following behavior under different lane changing stimulation was studied.The linear stability area is derived using linear analysis theory.The numerical simulation under different stimus with different competition coefficient is conducted to analysis the model rationality.(3)The LP-FVDM is proposed.The car-following process of the following vehicle on the current lant where the preeceeding vehicle lane pass from is discussed: the space gain factor is introduced to reflect the enlargement of the follower's front space casued by the lateral deviation of the preceding car,and the LP-FVDM is proposed by considering the space gain factor.The linear stability analysis is operated and the numerical simulation is conducted under different scenarios with different drivers' space gain factor.(4)Parameter calibration and model validation,the effectiveness of models are validated with the NGSIM data.The basic paraments of FVDM is calibrated using GA algorithm firstly,and the driver's driving style is recognized by the membership function of the time headway and the minimum time-headway,and the LM-FVDM and the LP-FVDM are tested.Above all,this paper proposed two car-following models considering preceding vehicle's lane-changing process.The numerical results show that the model can reasonably reflect the driver's velocity adjustment process under different LM and LP conditions.The NGSIM data experiments show that the proposed models are more close to the measured data than the FVDM under different scenarios,where the velocity error and position error of the proposed model are significantly lower than that of FVDM,which can reflects the speed change process more effectively in real traffic.