Research On Lane Keeping And Stability Of Electric Vehicles Under Human-manchine Cooperation

The lane keeping system of electric vehicles is of great significance for improving traffic safety.However,the current autonomous driving technology is not yet fully mature,and a fully autonomous lane keeping system is not yet able to adapt to the complex traffic environment.The collaborative control of the vehicle by the driver and the auxiliary control system and the lane keeping and stability control of the electric vehicle are the research hotspots in the field of vehicle intelligence and electrification.This thesis regards the 4WD electric vehicle as the subject of study,and studies the lane keeping and stability control of electric vehicles under human-machine coordination.First of all,the vehicle model of an electric car is established based on CarSim.Build the Dugoff tire model,motor model and PID speed controller in Simulink.At last the correctness of the model is proved by co-simulation.Aiming at the continuous human-machine cooperative co-driving mode,the driver and the auxiliary control system are always in the control loop,which causes the driver to continuously overcome the torque exerted by the auxiliary system during lane changes,which is likely to cause the problem of human-machine conflict.This paper proposes a vehicle auxiliary control system in which the vehicle control authority is switched between the driver and the control system.Avoid conflicts between man and machine during lane change.Based on the recognition of drivers’ lane changing intention,a TLC decision model is established.The TLC threshold is dynamically selected according to the driver’s reaction time and vehicle lateral stability requirements,so that the decision-making mechanism can automatically adjust the triggering time of lane departure warning and active steering intervention to avoid premature intervention of the control system and cause human-machine conflict.At the same time,based on the single point preview method,the lane keeping deviation model is established as the vehicle perception link.CarSim-Simulink co-simulation is used to verify the effectiveness of lane keeping decision.In lane keeping control,this paper proposes a control algorithm that combines lane tracking and stability control.In the stability control,for the sake of solving the problem that the β and μ are difficult to obtain their information,this paper designs a parameter observer based on EKF.On this basis,a lane keeping control system with a two-layer control structure is designed.Based on the MPC theory,a lane keeping and stability control system is designed.The function of the bottom controller is the realization of the yaw moment.According to the tire attachment ellipse,the torque is optimally distributed to realize the yaw moment.Finally,the co-simulation platform was built to simulate the effectiveness of the lane keeping driving assistance system with straight roads and curves as examples.In the simulation experiment,the lateral displacement deviation is kept within the acceptable range of ± 0.002 m in the straight lane and ± 0.2m in the curve.At the same time,slip angle can be effectively controlled within1°.The simulation shows that the established LKA system has good performance.At the same time,it can meet the stability requirements of the vehicle when driving on the road with low road adhesion coefficient.