V2X-Based Autonomous Driving Control At Intersections And Pedestrian Collision Warning

Intelligent Transportation System(ITS)has always been regarded as a popular research interest in the field of engineering applications.And autonomous driving and Internet of Vehicles(Io V)technology,which have been widespread attentioned and invested in recent years,have further promoted the development of ITS.In the future traffic scenes,some special scenes at present,such as intersections,will still be the key to frequent traffic accidents and low efficiency.Currently many researches are devoted to proposing reasonable solutions to the problems of driving safety and traffic efficiency of the future sence that autonomous vehicles cross intersections,as well as traffic collisions between vehicles and Vulnerable Road Users(VRUs)such as pedestrians around.The current researches on the process of autonomous driving crossing intersections are mostly suitable for ideal scenarios with simple traffic conditions,which are far from the actual situations,and usually introduce heavy communication and computing loads.In the meanwhile,the collision detection between vehicles and VRUs,most researches employ the methods like image processing to detect targets,which cannot handle the situation of obscured vision.In addition,most of them have proved the validity of the system for single target,however,lack of the performance tests for multi-targets.Based on the background above,this thesis focuses on intersections in future traffic scenarios and conducts the in-depth study on achieving the passing process of autonomous vehicles safely and efficiently under complex traffic conditions.Furthermore,a full-scene collision detection and warning system utilizing Io V communication for multiple VRU targets is also proposed to protect VRU safety.The main work of the thesis includes:(1)The research of trajectory planning strategy and control implementation during the process of autonomous vehicles traveling through intersections.First of all,the communication task,information format and content of each terminal from V2 X communication have been formulated in line with the principle of minimizing the communication load.Secondly,the trajectory planning strategy of autonomous vehicles employing the features of platoons aiming at the single-lane scene of the intersection is designed and introduced in detail,which is based on the idea of time delay and proved to achieve the transition process of vehicle states.And the lane selection strategy for the situation of multi-lane in the same direction is further proposed to ensure the shortest passing time required for vehicles.In addition,a model based on Model Predictive Control(MPC)is also established to perform vehicle control.Finally,the simulation results on MATLAB/Simulink prove the rationality and safety of the proposed strategy and validity of the control model.(2)Design the collision warning system between VRUs and vehicles.Firstly,the format and content of the transmission information are formulated under the premise of minimizing the communication load.Secondly,two Kalman filter-based models are proposed according to the different motion states of the target to reduce the error of GPS positioning information so as to improve the accuracy of the warning results.Then,the principle of the warning algorithm is introduced in detail.Finally,the simulation results on VISSIM show the validity of the proposed algorithm for potential collision prediction,as well as the availability and instantaneity in multi-target scenes.In addition,the experiments of the filter model also demonstrate its validity on reducing the positioning errors.