Study On Optimization Of Key Road And Traffic Parameters In Driverless Conditions

Driverless vehicle based on artificial intelligence and information technology is developing rapidly and relevant theories have been studied deeply and extensively.However,limited researches have been carried out on comprehensive road and traffic parameters under driverless condition.Generally,only limited traffic parameters have been considered.Based on the analysis of road and traffic parameters and key elements of driverless condition,a model of mixed driverless vehicles and manned vehicles was constructed by optimizing the existing car following model and lane changing model.Algorithm control and simulation were realized by VISSIM with the control of Python.The influence of different driverless vehicles’ proportions on multiple traffic parameters was analyzed,and the suggested values for related design speed under different driverless vehicles’ proportions were proposed through multi-objective optimization.Firstly,with multiple front vehicle speed introduced,MVCM car following model was established by optimizing OVCM model.Linear stability analysis was conducted,and disturbance propagation of the model was compared with classical FVD model and MHOVA model.Results showed that 1)increase of front vehicle number could effectively improve the stability of traffic flow,and the influence was significant enough when the front vehicle number was 4;2)MVCM model had better resistance to system disturbance than FVD model;3)compared with MHOVA model,MVCM model can obtain equal stability while considering less front vehicles.Secondly,lane changing model was established based on lane changing rule,lane changing path,speed planning and lane changing probability.This model was compared with Inter-Vehicle Communication model to analyze the relationship between traffic volume and density.Results showed that the relationship between traffic volume and density of the two models was very close,but speed curve of this model can reach stable earlier,and its speed fluctuation was smaller than that of Inter-Vehicle Communication model.Thirdly,according to relevant standards,a two-way four-lane urban arterial road model was established in VISSIM.The mode of mixed driverless vehicles and manned vehicles was realized by calling VISSIM COM interface through Python and the simulation strategy was established.Finally,the influence of different driverless vehicle proportions on multiple traffic parameters was analyzed.Results showed that 1)increase of driverless vehicle proportion improved the average vehicle speed and reduced the average delay time and average travel time.The influence was most significant when the proportion was between 0.4 and 0.6;2)average headway increased with the increase of driverless vehicle proportion when the traffic flow was relatively freer,and lane changing frequency decreased.Besides,a multi-objective optimization model based on decreasing travel time,increasing headway and reducing lane changing frequency was established.The multi-objective optimization problem was solved through NSGA-Ⅱ algorithm to obtain proposed values of design speed under different driverless vehicle proportion.This research realized the establishment of mixed driverless vehicles and manned vehicles model and established the traffic simulation platform.Analysis and evaluation methods of traffic flow parameters(delay,headway,lane changing frequency,etc.)and design speed were formed.This research provides a reference for the optimization of road and traffic design parameters for driverless conditions.