Research On The Path Planning And Tracking Of Intelligent Vehicle In Expressway Off-ramp Area

The intelligent vehicles promote the safty of road traffic.Currently,many Advanced Driver Assistant Systems on intelligent vehicles have been used widely.However,these functions are only applicable to simple conditions.The number of vehicles and express highway mileage are increasing with urbanization.Due to the heavy traffic flows and the different driving habits,the conditions of expressway off-ramp has changed negatively,leading to traffic accidents frequently.Therefore,Advanced Driver Assistant Systems may no longer be appropriate in the current complex situations,and certain high standards are required for intelligent vehicles.The main purpose of this paper is to study the path planning method and motion control strategy of intelligent vehicle for the scene of urban expressway ramp,so as to improve the safety of intelligent vehicle driving in the off-ramp area.Firstly,the artificial potential field model is improved to realize the path planning of the intelligent vehicle in the off-ramp area,and the repulsive potential field coefficient is optimized by using BP neural network model to improve the adaptability of the path planning model.Then,the PID velocity controller and the Model Predictive Control(MPC)and Stanley lateral controllers are built,thus the tracking effects of the controllers will be compared and analyzed.Finally,the urban expressway ramp scene,which based on the constructed joint simulation platform is established to carry out joint simulation and analyze rerfication of effectiveness of the control model through parameters such as tracking tracking trajectory,front wheel angle and vehicle body yaw angle.The research content is as follows:Firstly,using the improved artificial potential field model and BP neural network,the path planning of the ramp diversion area is completed.on the basis of artificial potential field method,the dynamic obstacle repulsion model,road model and gravitational model conforming to the off-ramp area are established,and a comprehensive constraint model of the artificial potential field is synthesized.The BP neural network is used to solve the problem of repulsion coefficient selection under different working conditions,realizing the path planning of urban expressway ramp under different working conditions.The effectiveness of the simulation model underpinned by the curvature of the front wheel is proved.Secondly,A intelligent vehicles motion control model is built,and the influence of different boundary conditions on the lateral controller is analyzed by tracking the double-shifting trajectory.Building on the PID control algorithm in the longitudinal direction,the vehicle speed controller is designed to ensure the effective tracking of the planned path during the driving process of the intelligent vehicle.In addition,the Stanley controller and the model predictive controller are constructed respectively using the two degree of freedom kinematic model and the three degree of freedom dynamic model as the foundation.The effect of different road adhesion coefficient and vehicle speed on the control of MPC and Stanley controller is also discussed.Thirdly,Car Sim vehicle dynamics software,Pre Scan scenario simulation software and the vehicle motion control model are the bases of the joint simulation platform built to set up four simulation conditions in the expressway on off-ramp area and complete the the validity verification of the intelligent vehicle motion control under different conditions.In conclusion,the improved artificial potential field method can plan an ideal driving path in the off-ramp area.The effect of road adhesion coefficient and vehicle speed on MPC transverse controller is less than that of Stanley controller.Co-simulation results show that the path planning algorithm and motion control algorithm proposed in this paper have a good controllability and adaptability.