Research On The Interaction Of Autonomous Vehicle On Bicyclists’ Behavior At Traffic Signal Intersections

On urban roads,non motorized vehicles,like their peers,belong to the disadvantaged side of road use,especially at T-junctions where left turns by non motorized vehicles are more likely to cause traffic accidents.Moreover,the current promotion of bicycle travel and the promotion of shared bicycles in China indicate that there will be a large number of cyclists in the future.For autonomous vehicle,due to the lack of human drivers,their interaction with non motor vehicles has become the focus of attention.The current research on the interaction between autonomous vehicle and pedestrians has proposed an external human machine interface(e HMI),which can interact with other road users by presenting information such as vehicle intentions.At the same time,vehicles can also express vehicle intentions through implicit kinematics cables,Currently,the activation sequence between explicit information(e HMI)and implicit information(vehicle kinematics clues)of autonomous vehicle is not clearly defined.Based on the above background,the purpose of this study is to explore the impact of the activation sequence of vehicle explicit and implicit information on cyclists’ behavior decisions when a single autonomous vehicle interacts with a single bicycle at a T-junction,and the impact of changes in pedestrian behavior on bicycles when an autonomous vehicle interacts with pedestrians and bicycles at the same time.This paper first designed two kinds of explicit and implicit information activation sequences and two kinds of activation intervals,and used the behind the scenes simulation method to conduct field experiments.A total of eight participants were recruited to carry out the experiment,with the purpose of exploring the impact of the activation sequence and activation interval of explicit and implicit information of autonomous vehicle on the behavior decision-making of riders;Then,on the basis of this experiment,the research subject was added to a single autonomous vehicle to interact with pedestrians and bicycles at the same time,and two kinds of behaviors that determine pedestrians’ crossing paths were proposed,namely,the direction and starting position of pedestrians,to explore the impact of pedestrian behavior changes on cyclists’ behavior decisions when autonomous vehicle interact with pedestrians and bicycles at the same time at T-junctions,also using behind the scenes simulation method,A total of 8 participants were recruited for the experiment;Finally,a virtual scene was constructed using the virtual reality method to verify the results of the previous experiment.The virtual reality method was used to add dangerous scenes and assign actual crossing behavior actions to pedestrians,exploring the impact of pedestrian behavior changes on cyclists’ decision-making in this scenario.The results show that:(1)When the autonomous vehicle interacts with the bicycle,the vehicle first presents explicit information and the implicit information is activated 2 seconds later,that is,it starts to decelerate and brake.At this time,the interaction between the rider and the autonomous vehicle is the most efficient.(2)The starting position of pedestrians has a significant impact on the total cycling time of cyclists.When the starting position of pedestrians is behind the road line,the cycling time of cyclists is relatively shorter,while when the starting position of pedestrians is in front of the road line,it greatly increases the cycling time of cyclists.For pedestrians’ direction of origin,it does not have any impact on cyclists.It is suggested that autonomous vehicle standardize the activation sequence of explicit information and implicit information,increase the efficiency of crossing traffic,increase the amount of information conveyed by the explicit interface,help other groups to clarify vehicle intentions,and improve the interaction efficiency between other groups.