Study On Modified Car-following Model Based On Autonomous Driving Environment

With rapid development of the Internet communication technology,the autonomous driving technology and its theory have become one of the major researches of many scholars and enterprises at present.And the analysis of the car-following model among autonomous cars is the basic job of this technology.Because the technology revolution should be adapted to the theoretical innovation simultaneously,this paper researches on the car-following model which is based on autonomous driving environment and builds a multi-information fusion car-following model(MI-CF)referring to the model under the traditional transportation environment.Moreover,the difference between the traditional model and the modified model in this paper is analyzed contrastively through the method of numerical simulation experiment,which verifies that the MI-CF model has higher stability as to demonstrate the characteristics of the car-following behaviors under the autonomous driving environment more accurately.First,this paper detailedly explains the research status of autonomous driving technology and the development of the car-following model.Besides,it introduces several kinds of classic traditional car-following models with comparing their advantages and disadvantages,finding that the traditional model mostly only consider the move status relationship between the follower car and the leader car without discussing the change under multiple information.The car-following behavior is more greatly affected by uncontrollable factors such as the characteristics of drivers,It is unable to accurately describe car-following behavior under autonomous driving environment.And then it analyzes the characteristics of both the car-following behaviors in autonomous traffic flow and the information interaction among cars based on autonomous car driving environment.Then,refer to the form of the modified velocity models,with the overall consideration of vehicle co-optimization velocity,velocity difference of more than one vehicle and space headway of more than one vehicle and other factors,this paper builds a multi-information fusion car-following model based on autonomous driving environment,Moreover,the parameters of the model are calibrated using the NGSIM data set of the Next Generation SIMulation(NGSIM)project supported by the Federal Highway Administration of the United States and the genetic algorithm.At last,this paper solves the stability conditions of the MI-CF model and make the numerical simulation experiment.Compares the model with traditional Optimal Velocity(OV)model and Full Velocity Difference(FVD)model,it is concluded that the MI-CF model in this paper has a larger stability region.After the numerical simulation experiment using virtual circular road method,comparing and analyzing the FVD model and the MI-CF model in this paper,the experimental results are in good agreement with the theoretical analysis results.The MI-CF model can make the vehicle respond to the change of the moving state of the vehicle in front faster and restore to the stable state in time.At the same time,it can maintain a more appropriate safety distance with the vehicle in front,which not only improves the efficiency of the fleet passing,but also ensures the safety of driving.Then sensitivity analysis of model parameters is made to explore the influence of some parameters on the stability of the model.This study can describe the car-following behavior of driverless traffic flow more accurately,and has certain theoretical significance for the future development of driverless technology.