Research On Design And Experiment Of Intelligent Vehicle Lane Departure Assistance System

With the upsurge of autonomous driving,advanced assisted driving system(ADAS)technology as a springboard for autonomous driving technology is becoming a hot research topic.Lane departure assistance system,one of the key technologies of advanced assisted driving system(ADAS),can effectively reduce traffic accidents caused by driver fatigue and distraction.In this dissertation,the intelligent vehicle lane departure assistance system is studied,the main research content is as follows.Firstly,this dissertation investigates the research content of today’s lane departure auxiliary system,and puts forward several technical requirements and system framework.The driver state intention recognition module is designed for the possible man-machine conflict.By studying the optimal presighting control theory,the ideal value of yaw velocity is deduced under the assumption of constant yaw velocity.The EPS of actuator was analyzed and the model was deduced.Secondly,based on the analysis of the current several early warning algorithms,considering the needs on space and time for early warning,a joint early warning algorithm based on TLC and FOD algorithm is proposed.In order to adapt to the driving habits of different drivers,fuzzy control is designed to optimize the joint early warning algorithm.Simulation experiments in MATLAB/Simulink and Carsim software environment verify the effectiveness of the early warning algorithm.Thirdly,a model predictive control(MPC)algorithm for lane keeping control is proposed to solve the problem that the tire side deflection is too large and the vehicle becomes unstable when the vehicle is under low cornering or road adhesion coefficient.The vehicle lateral position,yaw Angle,yaw velocity and yaw Angle of the center of mass were taken as the state variables and the front wheel rotation Angle as the control variables.Based on the two-degree-of-freedom vehicle model,the prediction model of the algorithm was established to track the ideal yaw Angle under the optimal preview control theory,and the constraint of yaw Angle of the tire was designed to ensure the the vehicle stable.MATLAB/Simulink and Carsim platform were used for the joint simulation experiment,which verified the effectiveness of the MPC algorithm in this dissertation.Finally,the construction and debugging of the real vehicle test platform were carried out.The software and hardware of the test vehicle were installed and calibrated from the four aspects of perception,decision-making,execution and communication.The real vehicle test of lane departure warning and lane keeping was carried out based on the test vehicle.