Study On Lane Change Of Autonomous Vehicle In Continuous Traffic Flow Environment

With the development of science and technology such as artificial intelligence,autonomous vehicles have gradually appeared in the public’s view.At present,some progress has been made in autonomous driving technology,and China has put forward the beautiful vision of gradually becoming a powerful nation of intelligent car by 2035-2050.It is not difficult to imagine that in the future,autonomous vehicles will become a new way to solve traffic congestion problems due to their high efficiency of information interaction and high uniformity of driving technology.At the same time,in the process of its development,there is bound to be a mixed stage of manual and automatic driving vehicles.Nowadays,major cities have built urban expressways to alleviate the traffic pressure of main roads,but congestion will still occur in the morning and evening peak.In this continuous traffic flow state,the structural factors of the road and the inappropriate driving behavior of drivers are often the main causes of traffic congestion.This includes the driver’s forced lane change due to the structure of the expressway.Because of its strong purpose,forced lane change often has a greater impact on the traffic flow than free lane change.Therefore,it is necessary to study the forced lane changing behavior of autonomous vehicles under continuous traffic flow,and consider two driving environments: fully autonomous driving and manual automatic hybrid driving.In the fully autonomous driving environment,considering the strong cooperation between autonomous vehicles,taking the speed of lane changing vehicles and vehicles behind the target lane as the main indicator to ensure the right of way of vehicles on the main road,and taking the successful lane changing and the distance from the off ramp as the secondary indicators to ensure the lane changing demand of lane changing vehicles,a two lane income calculation model is established,and a decision with higher income value is obtained through comparison,Guide vehicles in the target lane to change lanes cooperatively.In the artificial automatic hybrid driving environment,the interaction process between autonomous vehicles and human driven vehicles is described according to the idea of game theory.Taking the headway as the main indicator to ensure the safety of lane changing,and the vehicle speed and lane changing as the secondary indicators,a lane changing game model is established.The benefits of the two vehicles in the lane changing process are calculated respectively,and the Nash equilibrium method is used to solve the model to obtain the optimal strategy.At the same time,considering the influence of different driver styles on the final strategy,the model parameters are modified by driving style recognition.Finally,VISSIM and MATLAB are used to simulate the two driving environments,and compared with the existing lane change models to obtain the lane change position,speed,delay and other data to verify the effectiveness of the model proposed in this paper.The results show that the two-lane revenue model can generally make the lane changing vehicles change to the outermost lane when they are 600 m away from the off ramp,which is about 300 m earlier than the existing model;When the vehicle is in the second lane and 1200 m away from the off ramp,the model will not get the lane change decision;The vehicle lane changing speed is 10km/h higher than the existing model,and will not affect the overall traffic flow efficiency.The delay of lane changing game model considering driver’s style is3-5s less than that of ordinary game model;For aggressive drivers,the initial distance between the lane changing vehicle and the vehicle behind the target lane is within 80 m,and the model will get the decision not to change lanes.For conservative drivers,the distance is50 m,which has good safety.