Research And Application Of Conditional Autonomous Vehicle Behavior Modeling

With the deep integration of the automotive industry with the Internet of things,artificial intelligence and other new generation technologies in recent years,the conditions of conditional autonomous vehicles have been put into use one after another,providing new ideas for the improvement of traffic conditions.Conditional autonomous vehicles are vehicles that can perform most of the operations independently from the driver’s control,and are significantly different from manually driven vehicles in terms of information acquisition,behavioral decision making,and control execution.Therefore,in order to better characterize the behavior of conditional autonomous vehicles,this thesis investigates their following behavior and lane changing behavior.The main innovations and work of this thesis are as follows:(1)A model of following behavior of a conditional autonomous vehicle in a singlelane environment is proposed.The following model based on the information of multiple vehicles in front is constructed for the characteristics of the conditional autonomous driving vehicle to accurately acquire vehicle information in a larger range.The model makes up for the shortage that the artificially driven vehicle is only influenced by the driving condition of one following vehicle in the immediate vicinity,and fully considers the optimized speed and safety distance of multiple following vehicles in front of it to realize its own driving speed and acceleration adjustment.In this thesis,simulation experiments on driving behavior are conducted using the meta-cellular automata model,and the experimental results show that the number of monitorable vehicles in front has different degrees of influence on the stability of traffic flow,which verifies the effect of conditional autonomous vehicles on the improvement of road traffic conditions.(2)A lane change behavior model for conditional autonomous vehicles in a multilane environment is proposed.Aiming at the situation that conditional autonomous vehicles make lane changes in a multi-lane environment because they cannot meet the needs of driving at desired speeds and driving in open spaces,this thesis introduces a three-lane neighbor model,improves the STCA model for the formulation of lane change rules,and constructs a lane change model applicable to conditional autonomous vehicles.The model analyzes the optimal average speed of adjacent lanes,selects the most suitable lane to continue driving and avoids invalid lane change,and compares the difference of traffic flow conditions under different ratios of conditional autonomous vehicles.The experimental results show that conditional autonomous vehicles have obvious effects on improving traffic congestion and increasing road utilization.(3)A microscopic traffic simulation system including conditional autonomous vehicles is designed and implemented.The system supports the simulation of conditional autonomous driving vehicle behavior,and in addition,it also has the functions of simulation road network drawing,real road network transformation,simulation model selection,simulation data generation,etc.to realize the visualization of simulation process and result data display.Through the test,the system has good functional realization and usability.The work in this thesis involves theoretical research and engineering applications.The modeling study of the following behavior and lane changing behavior of the conditional autonomous vehicles can describe the differences between the behavior of the vehicles and that of the artificially driven vehicles,and the simulation system can realize the simulation,verification and analysis of the behavior of the conditional autonomous vehicles,and the research results of this thesis can be used for the improvement of the stability of traffic flow and road utilization by the conditional autonomous vehicles.