Design And Realization Of The Connected And Autonomous Vehicle Lane-changing Strategies In Mixed Traffic

The continuous progress of vehicle-vehicle/vehicle-infrastructure(V2V/V2I)communication and vehicle autonomous driving technologies have greatly promoted the development of connected and autonomous vehicles(CAVs)in recent years.At the same time,under the guidance of relevant policies,the open road test of CAVs has been further promoted.Therefore,mixed traffic flow scenarios including CAVs,self-driving vehicles and human-driven vehicles(HVs)will be an inevitable trend for a long time in the near future.Due to the dynamics and random natures of human driver behaviors,mixed traffic flow scenarios place higher demands on the collaboration between CAVs and HVs.As a typical application scenario of V2 V collaboration,lane changing has great value in both theoretical research and industrial application.Existing studies of CAV lane-changing strategies focus on the lane-changing behavior of a single CAV,and there is still no study focus on the dynamic lane-changing strategies of CAVs and the collaborative lane-changing strategies among multiple CAVs.Based on the existing CAV technologies,this paper focuses on the research gaps in the dynamics of the human driver behaviors in mixed traffic flow scenarios,the impact of lane-changing behavior on traffic flow,and the collaborative mandatory lane changing among CAVs,and proposes relevant lane-changing models.Comprehensive experiments have been conducted to verify the proposed lane-changing models in various traffic scenarios.Specific research contents are listed as follows.(1)Propose a CAV lane-changing model which considering the dynamic behavior characteristics of HVs.Four key steps including linearized car-following model,lane-changing decision,dynamic lane-changing trajectory generation,and high-precision trajectory tracking work tightly to handle the rapid acceleration/deceleration,cooperative/uncooperative,and simultaneous lane changes of HVs that may occur during the lane-changing process of CAVs.And innovatively propose a response mechanism for lane-changing failures,and elaborated the countermeasures for lane-changing failures in a dynamic environment.Design and implement experiments in four typical traffic scenarios.The experimental results confirm the dynamics and applicability of the proposed lane-changing model.(2)Propose a multi-CAV cooperative lane-changing model toward mixed traffic scenarios.On the basis of constructing the vehicle longitudinal kinematic model,the lane-changing decision is made by predicting the CAVs' trajectories,and the cooperative lane-changing trajectory planning is carried out with the aim of minimizing the impact on the platoon on the target lane.97461 traffic scenarios are designed to verify the effectiveness of the proposed cooperative lane-changing model in the experiments.And traditional lane-changing model is used as comparison.Experimental results show that the proposed cooperative lane-changing model is capable of improving the lane-changing success rate by 36.7% compared with the traditional lane-changing model,and greatly reduce the impact on the platoon on the target lane.(3)Propose a cooperative mandatory lane-changing model for multi-CAVs in mixed traffic scenarios.In order to better cope with the mandatory lane-changing scenarios such as vehicles driving out of the highway,reducing the number of lanes ahead,and avoiding obstacles on this lane.This study is based on collaboration between CAVs,and calculates the lateral acceleration of the lane change through the cubic curve to divide the road area to free lane-changing area,mandatory lane-changing area,and mandatory lane-changing execution area.In the free lane-changing area,the lane change operation is performed on the premise that it does not affect the normal operation of the vehicles behind on the target lane;in the mandatory lane-changing area,multiple vehicles cooperate to coordinate the lane change trajectory planning to ensure the safety of the mandatory lane change task.The simulation experiments including two typical traffic scenarios and 735 traffic conditions confirm the applicability and reliability of the proposed mandatory lane-changing model.(4)Design and develop a CAV.The six parts of CAV key technologies,including V2V/V2 I communication,localization and navigation,environment perception,decision-making and planning,trajectory tracking,and vehicle drive-by-wire control are introduced in detail.In addition,combined with the research and development of the CAV platform and the test site,a three-degree-of-freedom dynamic kinematic model of the entire vehicle is established,and a high-precision trajectory tracking algorithm based on model predictive control is improved,which enhances the trajectory tracking accuracy.The operation efficiency ensures the real-time performance of the algorithm.(5)In order to conduct a more comprehensive test and application of the proposed lane-changing model,a field experiment is carried out at the test site to verify the proposed model.Through setting up a variety of different test scenarios,a comprehensive test and verification of the CAV platform and the lane-changing model developed are conducted.Experimental results show that the proposed lane-changing model can adapt to a variety of different traffic scenarios and the dynamic behavior of different HVs,and ensure the safety and comfort during the lane-changing process.Compared with the human drivers' lane-changing behaviors in the same scene,the proposed lane-changing model achieves a safer and more comfortable lane-changing operation without reducing the lane-changing efficiency.The paper summarizes the advantages and disadvantages of the multiple lane-changing models proposed in the study based on the experimental results,and also puts forward the prospect of the next step of the research work.