Research On Lane Keeping And Lane Change Behavior Decision Method Of Intelligent Vehicle

Since the last century,the number of traffic accidents in my country has been increasing,and the proportion of traffic accidents caused by vehicles changing lanes or overtaking has reached as high as 60%.In order to improve traffic conditions and promote the development of the autonomous driving industry,this topic is based on high-speed scenarios and conducts research on the whole process of lane keeping and lane change behavior decision-making.The research content is as follows:(1)Aiming at the limitations of existing algorithms for lane keeping and lane change behavior decision-making,a new idea of fusing the driving risk field and driver dissatisfaction is proposed.Establish the potential energy field and kinetic energy field model of the relevant traffic environment for the driving risk field model,including the potential energy field of the lane line,the kinetic energy field of the stationary vehicle,and the kinetic energy field model of the moving vehicle,and perform simulation verification on the model to explain some of its existence Disadvantages.Establish a driver dissatisfaction model based on the actual drivers’ emotional changes during driving,and establish a driver’s dissatisfaction model based on the driver’s behavioral characteristics of pursuing the target speed and the relationship between the target speed and the background vehicle speed.Simulate and explain its shortcomings.Then the two models are fused to overcome the limitations of the respective algorithms for lane keeping and lane change behavior decision-making.(2)Establish a decision-making model for lane keeping and lane change behavior.After fusing the driving risk field model and the driver’s dissatisfaction model,the entire process of lane keeping and lane change decision-making is divided,and lane keeping and decision-making algorithms are designed.Lane keeping algorithm design includes following control,lane departure judgment,and lane keeping control.The design of lane change decision algorithm includes lane change intention generation,target lane selection,lane change feasibility analysis and determination of lane change time threshold.In the above algorithm design,lane keeping control is constrained by the potential energy field formed by the lane line;lane change feasibility analysis is judged by the comprehensive driving risk field formed by the surrounding traffic environment;the intention of lane change and the target lane The choice is judged by the drivers’ dissatisfaction;then the simulated driving experiment is carried out through the simulation of the driver combined with the Prescan simulation software and the driving data of the real driver is collected,and the acquired data is quantitatively analyzed to determine the threshold of the specific time of the lane change.(3)Carry out simulation realization on lane keeping and lane change decision algorithm.Use the means of driver dissatisfaction to verify the data collected in the simulation driving experiment,and analyze the following four typical driving scenarios:a background car scene in front of the main car,a background car scene in front of the target lane,and a background car scene in front of the target lane.There are background car scenes before and after the target lane and the background car scene,and then the simulation experiment is carried out on the above scenes.The experimental results show that the fusion method of the driving risk field model and the driver dissatisfaction model used in this project can effectively solve the problem of smart car lane keeping and lane change behavior decision-making.